8481 articles  [english version]

inria-00070434, version 1

Random thresholds for linear model selection

Marc Lavielle 1, Carenne Ludeña

N° RR-5572 (2005)

Résumé : A method is introduced to estimate the number of significant coefficients in non ordered model selection problems. The method is based on a convenient random centering of the partial sums of the ordered observations. Based on $L-$statistics methods we show consistency of the proposed estimator. An extension to unknown parametric distributions is considered. The method is then applied to a regression model and interpreted as a random threshold procedure. Simulated examples are included to show the accuracy of the estimator.

  • 1 :  SELECT (INRIA Futurs)
  • INRIA – Université Paris XI - Paris Sud
  • Domaine : Informatique/Autre
  • Mots-clés : ADAPTIVE ESTIMATION / LINEAR MODEL SELECTION / HARD THRESHOLDING / RANDOM THRESHOLDING / $L$ STATISTICS
  • Référence interne : RR-5572
 
  • inria-00070434, version 1
  • http://hal.inria.fr/inria-00070434
  • oai:hal.inria.fr:inria-00070434
  • Contributeur : Rapport De Recherche Inria <>
  • Soumis le : Vendredi 19 Mai 2006, 20:28:46
  • Dernière modification le : Mercredi 31 Mai 2006, 14:24:23