您的位置:山东大学 -> 科技期刊社 -> 《山东大学学报(理学版)》

J4 ›› 2011, Vol. 46 ›› Issue (1): 97-103.

• 数学 • 上一篇    下一篇

广义二项回归模型的影响诊断

赵为华1, 冯予2, 李泽安3   

  1. 1. 南通大学理学院, 江苏 南通 226007;  2.  南京理工大学理学院, 江苏 南京  210094;
    3. 南通大学计算机学院, 江苏 南通  226007
  • 收稿日期:2010-01-27 出版日期:2011-12-14 发布日期:2011-03-16
  • 作者简介:赵为华(1978-),男,讲师,博士在读,主要从事数理统计方面的研究。Email: zhao.wh@ntu.edu.cn
  • 基金资助:

    江苏省自然科学基金资助项目(SBK200920379);南通大学自然科学基金资助项目(08Z032)

Influence diagnostics for generalized binomial regression models

ZHAO Wei-hua1, FENG Yu2, LI Ze-an3   

  1. 1. School of Science, Nantong University, Nantong 226007, Jiangsu, China;
     2. School of Science, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China;
     3.  School of Computer, Nantong University,  Nantong  226007, Jiangsu, China
  • Received:2010-01-27 Online:2011-12-14 Published:2011-03-16

摘要:

本文系统研究了广义二项回归模型的统计诊断和影响分析方法, 给出了模型的参数估计和模型中散度参数检验的Score统计量,并基于数据删除模型和在3种局部扰动情形下,研究了模型影响分析问题,最后通过两个数值例子验证了本文方法的有效性。

关键词: 广义二项回归; 统计诊断; 影响分析;Score检验; 广义Cook距离

Abstract:

The diagnostics and influence analysis methods of the generalized binomial regression model were sudied. The estimation of the parameters and the Score test of the dispersion parameters about the model are given. And the influence analysis were sudied base on the case-deletion model and under three kinds of local influence. Finally, two numerical examples illustrateed the effectiveness of the method.

Key words: generalized binomial regression; diagnostics; influence analysis; Score test; generalized Cook distance

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!