数学物理学报(英文版) ›› 2000, Vol. 20 ›› Issue (4): 563-570.

• 论文 • 上一篇    下一篇

BAYESIAN LOCAL INFLUENCE ASSESSMENTS IN A GROWTH CURVE MODEL WITH GENERAL COVARIANCE STRUCTURE

 白鹏, 费宇   

  1. Department of Statistics, Yunnan University, Kunming 650091, China Institute of Applied Mathematics, Yunnan Province, Kunming 650091, China
  • 收稿日期:1998-06-04 修回日期:1999-09-28 出版日期:2000-06-15 发布日期:2000-06-15
  • 基金资助:

    Supported by the fund of the Yunnan Education Committee
    (NO.9941072)

BAYESIAN LOCAL INFLUENCE ASSESSMENTS IN A GROWTH CURVE MODEL WITH GENERAL COVARIANCE STRUCTURE

 BAI Feng, BI Yu   

  1. Department of Statistics, Yunnan University, Kunming 650091, China Institute of Applied Mathematics, Yunnan Province, Kunming 650091, China
  • Received:1998-06-04 Revised:1999-09-28 Online:2000-06-15 Published:2000-06-15
  • Supported by:

    Supported by the fund of the Yunnan Education Committee
    (NO.9941072)

摘要:

The objective of this paper is to present a Bayesian approach based on Kullback-
Leibler divergence for assessing local influence in a growth curve model with general co-
variance structure. Under certain prior distribution assumption, the Kullback-Leibler di-
vergence is used to measure the influence of some minor perturbation on the posterior
distribution of unknown parameter. This leads to the diagnostic statistic for detecting
which response is locally influential. As an application, the common covariance-weighted
perturbation scheme is thoroughly considered.

关键词: Growth curve model, prior and posterior distribution, Kullback-Leibler di-
vergence, Bayesian !-model, curvature

Abstract:

The objective of this paper is to present a Bayesian approach based on Kullback-
Leibler divergence for assessing local influence in a growth curve model with general co-
variance structure. Under certain prior distribution assumption, the Kullback-Leibler di-
vergence is used to measure the influence of some minor perturbation on the posterior
distribution of unknown parameter. This leads to the diagnostic statistic for detecting
which response is locally influential. As an application, the common covariance-weighted
perturbation scheme is thoroughly considered.

Key words: Growth curve model, prior and posterior distribution, Kullback-Leibler di-
vergence, Bayesian !-model, curvature

中图分类号: 

  • 62H