Generalized additive modelling of sample extremes

被引：203

作者：

Chavez-Demoulin, V

Davison, AC ^{[1
]}

机构：

[1] Swiss Fed Inst Technol, Sch Basic Sci, Inst Math, CH-1015 Lausanne, Switzerland

[2] ETH, Zurich, Switzerland

来源：

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS | 2005年 / 54卷

关键词：

bootstrap; generalized additive model; generalized pareto distribution; natural cubic spline; North Atlantic oscillation; parameter orthogonality; peaks over threshold; penalized likelihood; statistics of extremes; temperature data;

D O I：

10.1111/j.1467-9876.2005.00479.x

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

We describe smooth non-stationary generalized additive modelling for sample extremes, in which spline smoothers are incorporated into models for exceedances over high thresholds. Fitting is by maximum penalized likelihood estimation, with uncertainty assessed by using differences of deviances and bootstrap simulation. The approach is illustrated by using data on extreme winter temperatures in the Swiss Alps, analysis of which shows strong influence of the north Atlantic oscillation. Benefits of the new approach are flexible and appropriate modelling of extremes, more realistic assessment of estimation uncertainty and the accommodation of complex dependence patterns.

引用

页码：207 / 222

页数：16

共 35 条

[1]

AKAIKE H, 1973, INT S INFORMATION TH, V2, P267

[2]

Anderson C.W., 1990, J R STAT SOC B, V52, P425

[3]

[Anonymous], 2003, Statistical Models

[4]

[Anonymous], 2002, ANAL LONGITUDINAL DA