Separate inference induced from the structure of a statistical model
Project/Area Number  08680340 
Research Category 
GrantinAid for Scientific Research (C)

Section  一般 
Research Field 
Statistical science

Research Institution  The Institute of Statistical Mathematics 
Principal Investigator 
YANAGIMOTO Takemi The Institute of Statistical Mathematics, Department of Interdisciplinary Statistics, Professor, 領域統計研究系, 教授 (40000195)

Project Fiscal Year 
1996 – 1997

Project Status 
Completed(Fiscal Year 1997)

Budget Amount *help 
¥1,700,000 (Direct Cost : ¥1,700,000)
Fiscal Year 1997 : ¥600,000 (Direct Cost : ¥600,000)
Fiscal Year 1996 : ¥1,100,000 (Direct Cost : ¥1,100,000)

Keywords  Conditional MLE / GEE / Marginal Likelihood / Empirical Bays method / MLE / 条件付MLE / 一般化推定方程式 / 周辺尤度 / 経験ベイズ法 / 最尤法 / 直交性 / 推定関数の不変性 / 多次元母数 / スタイン推定量 / 対数線型モデル 
Research Abstract 
1) It becomes clear that a useful model in practical applications involves in more parameters (parameters of highdemensions) than it is previously believed. A lowdimensional parameter is unble to describe a practical situation. Here we should note that the crude MLE behaves unfavorably in estimating a highdimensional parameter. My previous efforts went to modifying the MLE through the conditional MLE,the estimating equation. 2) The present research aimed at introducing a new family of distributions named the ruled exponential distributions. Then analytical properties of the family and their applications were discussed. A simple estimating function is investigated in relation with a family of distributions. 3) More explicitly, the MLE based on a separate likelihood, mainly in the ruled exponential family, was pursued. Further, the present research extended to an obvious application to the estimating function and also to a modification of the family. 4) A challenging topic is to unify the generalized regression model and the Stein type estimator. The present approach looks promising. We are making efforts to contributing this important problem, though a bulk of difficulties arise.

Report
(4results)
Research Output
(11results)