文章基本信息

标题：On linear mixed models and their influence diagnostics applied to an actuarial problem
本地全文：下载
作者：Luis Gustavo Bastos Pinho ; Juvêncio Santos Nobre ; Sílvia Maria de Freitas 等
期刊名称：Chilean Journal of Statistics
印刷版ISSN：0718-7912
电子版ISSN：0718-7920
出版年度：2012
卷号：3
期号：1
页码：57-73
出版社：Chilean Statistical Society
摘要：In this paper, we motivate the use of linear mixed models and diagnostic analysis in practical actuarial problems. Linear mixed models are an alternative to traditional cred- ibility models. Frees et al. (1999) showed that some mixed models are equivalent to some widely used credibility models. The main advantage of linear mixed models is the use of diagnostic methods. These methods may help to improve the model choice and to iden- tify outliers or in°uential subjects which deserve better attention by the insurer. As an application example, the data set in Hachemeister (1975) is modeled by a linear mixed model. We can conclude that this approach is superior to the traditional credibility one since the former is more °exible and allows the use of diagnostic methods
关键词：Credibility models ￠ Hachemeister model ￠ Linear mixed models;￠ Diagnostics ￠ Local in°uence ￠ Residual analysis.