21787 articles – 15600 references  [version française]

hal-00461580, version 2

Mirror averaging with sparsity priors

Arnak Dalalyan () 1, Alexandre Tsybakov () 2

(2010-03-05)

  • 1:  Laboratoire d'Informatique Gaspard-Monge (LIGM)
  • http://igm.univ-mlv.fr/LIGM/
    Université Paris-Est Marne-la-Vallée (UPEMLV) – ESIEE – Ecole des Ponts ParisTech – Fédération de Recherche Bézout – CNRS : UMR8049 Université de Paris-Est - Marne-la-Vallée, Cité Descartes, Bâtiment Copernic, 5 bd Descartes, 77454 Marne-la-Vallée Cedex 2, Inst Gaspard Monge France
  • 2:  Laboratoire de Probabilités et Modèles Aléatoires (LPMA)
  • http://www.proba.jussieu.fr/
    CNRS : UMR7599 – Université Pierre et Marie Curie [UPMC] - Paris VI – Université Paris VII - Paris Diderot France
  • Available versions :  v1 (2010-03-05) v2 (2010-11-25) v3 (2012-07-27)
  • Bibliographic reference

    • Type of document: Documents without publication reference (Preprint)
    • Subject:
      Mathematics/Statistics
      Statistics/Statistics Theory
    • Title: Mirror averaging with sparsity priors
    • Abstract: We consider the problem of aggregating the elements of a (possibly infinite) dictionary for building a decision procedure, that aims at minimizing a given criterion. Along with the dictionary, an independent identically distributed training sample is available, on which the performance of a given procedure can be tested. In a fairly general set-up, we establish an oracle inequality for the Mirror Averaging aggregate based on any prior distribution. This oracle inequality is applied in the context of sparse coding for different problems of statistics and machine learning such as regression, density estimation and binary classification.
    • Fulltext language: English
    • Production date: 2010-03-05
    • Keyword(s): Mirror averaging – progressive mixture – sparsity – aggregation of estimators – oracle inequalities
    • ANR Project:
      Project Id PARCIMONIE

    Attached file list to this document: 

     
    • hal-00461580, version 2
    • oai:hal.archives-ouvertes.fr:hal-00461580
    • From: 
    • Submitted on: Thursday, 25 November 2010 07:58:31
    • Updated on: Friday, 26 November 2010 11:40:01