Spatial extreme quantile estimation using a weighted log-likelihood approach

Julie Carreau 1 Stephane Girard 2
2 MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, Institut National Polytechnique de Grenoble (INPG)
Abstract : We propose to estimate spatial extreme quantiles by a weighted log-likelihood approach. It is assumed that the conditional distribution of the variable of interest follows a generalized extreme-value distribution. The associated response surfaces are estimated thanks to the introduction of weights in the log-likelihood. These weights depend on the distance between the point of interest and the observations. The construction of a proper distance relies on the combination of a multidimensional scaling unfolding with a neural network regression. Our approach is illustrated both on simulated and real rainfall datasets.
Document type :
Journal articles
Journal de la Société Française de Statistique, Société Française de Statistique et Société Mathématique de France, 2011, 152 (3), pp.66-82


https://hal.archives-ouvertes.fr/hal-00761361
Contributor : Stephane Girard <>
Submitted on : Thursday, December 6, 2012 - 10:21:12 AM
Last modification on : Thursday, February 26, 2015 - 10:59:56 AM

File

sfds_jsfds_152_3_66-82.pdf
fileSource_public_greenPublisher

Identifiers

  • HAL Id : hal-00761361, version 1

Collections

Citation

Julie Carreau, Stephane Girard. Spatial extreme quantile estimation using a weighted log-likelihood approach. Journal de la Société Française de Statistique, Société Française de Statistique et Société Mathématique de France, 2011, 152 (3), pp.66-82. <hal-00761361>

Export

Share

Metrics

Consultation de
la notice

178

Téléchargement du document

84