Resumé
We consider bias-reduced estimation of the extreme value index in conditional Pareto-type models with random covariates when the response variable is subject to random right censoring. The bias-correction is obtained by fitting the extended Pareto distribution locally to the relative excesses over a high threshold using the maximum likelihood method. Convergence in probability and asymptotic normality of the estimators are established under suitable assumptions. The finite sample behaviour is illustrated with a simulation experiment and the method is applied to two real datasets.
| Originalsprog | Engelsk |
|---|---|
| Tidsskrift | Extremes |
| Vol/bind | 22 |
| Udgave nummer | 3 |
| Sider (fra-til) | 459-498 |
| Antal sider | 40 |
| ISSN | 1386-1999 |
| DOI | |
| Status | Udgivet - 15. sep. 2019 |
Fingeraftryk
Citer dette
}
Bias-corrected estimation for conditional Pareto-type distributions with random right censoring. / Goegebeur, Yuri; Qin, Jing; Guillou, Armelle.
I: Extremes, Bind 22, Nr. 3, 15.09.2019, s. 459-498.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › peer review
TY - JOUR
T1 - Bias-corrected estimation for conditional Pareto-type distributions with random right censoring
AU - Goegebeur, Yuri
AU - Qin, Jing
AU - Guillou, Armelle
PY - 2019/9/15
Y1 - 2019/9/15
N2 - We consider bias-reduced estimation of the extreme value index in conditional Pareto-type models with random covariates when the response variable is subject to random right censoring. The bias-correction is obtained by fitting the extended Pareto distribution locally to the relative excesses over a high threshold using the maximum likelihood method. Convergence in probability and asymptotic normality of the estimators are established under suitable assumptions. The finite sample behaviour is illustrated with a simulation experiment and the method is applied to two real datasets.
AB - We consider bias-reduced estimation of the extreme value index in conditional Pareto-type models with random covariates when the response variable is subject to random right censoring. The bias-correction is obtained by fitting the extended Pareto distribution locally to the relative excesses over a high threshold using the maximum likelihood method. Convergence in probability and asymptotic normality of the estimators are established under suitable assumptions. The finite sample behaviour is illustrated with a simulation experiment and the method is applied to two real datasets.
KW - 62G08
KW - 62G20
KW - 62G32
KW - Bias-reduction
KW - Local estimation
KW - Pareto-type model
KW - Random covariate
KW - Random right censoring
U2 - 10.1007/s10687-019-00341-7
DO - 10.1007/s10687-019-00341-7
M3 - Journal article
VL - 22
SP - 459
EP - 498
JO - Extremes
JF - Extremes
SN - 1386-1999
IS - 3
ER -