 |
|
|
|
| HAL : hal-00306219, version 3 |
| arXiv : 0807.4081 |
| DOI : 10.1214/09-EJS430 |
| Fiche détaillée |  Récupérer au format |
|
|
| Electronic Journal of Statistics 3 (2009) 678-711 |
|
|
| Versions disponibles : | v1 (25-07-2008) | v2 (10-06-2009) | v3 (13-07-2009) |
|
|
|
|
| Optimal weighting for false discovery rate control |
|
|
| Etienne Roquain 1Mark Van De Wiel 2, 3 |
|
|
| (10/07/2009) |
|
|
|
| How to weigh the Benjamini-Hochberg procedure? In the context of multiple hypothesis testing, we propose a new step-wise procedure that controls the false discovery rate (FDR) and we prove it to be more powerful than any weighted Benjamini-Hochberg procedure. Both finite-sample and asymptotic results are presented. Moreover, we illustrate good performance of our procedure in simulations and a genomics application. This work is particularly useful in the case of heterogeneous $p$-value distributions. |
|
|
|
|
|
|
|
|
|
|
|
| 1 : | Laboratoire de Probabilités et Modèles Aléatoires (LPMA) |
|
|
|
CNRS : UMR7599 – Université Paris VI - Pierre et Marie Curie – Université Paris VII - Paris Diderot |
|
|
| 2 : | Department of Mathematics [Amsterdam] |
|
|
|
Universiteit van Amsterdam |
|
|
| 3 : | Dep. of Epidemiology and Biostatistics |
|
|
|
Medical Center, Vrije Universiteit |
|
|
|
|
|
|
|
|
|
| Domaine | : | Mathématiques/Statistiques Statistiques/Théorie |
|
|
| False discovery rate – multiple testing – $p$-value weighting – power maximization. |
|
|
|
| Liste des fichiers attachés à ce document : | ||||||||||
|
||||||||||
|
|
|
| hal-00306219, version 3 | |
| http://hal.archives-ouvertes.fr/hal-00306219 | |
| oai:hal.archives-ouvertes.fr:hal-00306219 | |
| Contributeur : Etienne Roquain | |
| Soumis le : Lundi 13 Juillet 2009, 18:00:18 | |
| Dernière modification le : Lundi 13 Juillet 2009, 18:15:55 | |
