Ranking and empirical minimization of U-statistics - Université Paris Nanterre Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2006

Ranking and empirical minimization of U-statistics

Résumé

The problem of ranking/ordering instances, instead of simply classifying them, has recently gained much attention in machine learning. In this paper we formulate the ranking problem in a rigorous statistical framework. The goal is to learn a ranking rule for deciding, among two instances, which one is "better," with minimum ranking risk. Since the natural estimates of the risk are of the form of a U-statistic, results of the theory of U-processes are required for investigating the consistency of empirical risk minimizers. We establish in particular a tail inequality for degenerate U-processes, and apply it for showing that fast rates of convergence may be achieved under specific noise assumptions, just like in classification. Convex risk minimization methods are also studied.
Fichier principal
Vignette du fichier
ranking_submit.pdf (349.33 Ko) Télécharger le fichier

Dates et versions

hal-00020087 , version 1 (05-03-2006)

Identifiants

Citer

Stéphan Clémençon, Gábor Lugosi, Nicolas Vayatis. Ranking and empirical minimization of U-statistics. 2006. ⟨hal-00020087⟩
116 Consultations
96 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More