An AIC for Selecting Frailty Structures

• 발행기관: 한국자료분석학회
• 저자명: 하일도(Il Do Ha) 조건호(Geon-Ho Cho)
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.7 No.1, 41~47쪽, 전체 7쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2005.02.28

영문 초록

Frailty models with various structures such as shared, AR(1) or nested frailty structures have been developed, and are now widely used for analyzing multivariate survival data in the form of recurrent or multiple event times. For selecting a frailty structure in the(possibly) non-nested frailty models Ha and Cho(2003) used a new AIC(Akaike information criterion), based on hierarchical likelihood. However, they didn t provide an appropriate justification on this model-selection criterion. In this paper, we show how to construct reasonably this AIC from the HGLM(hierarchical generalized linear model, Lee and Nelder, 1996, 2001) setting and demonstrate via a numerical study that the AIC works well.

목차

1. Introduction

2. An AIC for Frailty Models

3. Numerical Study

4. Discussion

References

키워드

Akaike information criterion Frailty models Hierarchical likelihood Model selection Multivariate survival data.