Maximal deviations of incomplete U-statistics with applications to empirical risk sampling - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2013

Maximal deviations of incomplete U-statistics with applications to empirical risk sampling

Résumé

It is the goal of this paper to extend the Empirical Risk Minimization (ERM) paradigm, from a practical perspective, to the situation where a natural estimate of the risk is of the form of a K-sample U-statistics, as it is the case in the K-partite ranking problem for instance. Indeed, the numerical computation of the empirical risk is hardly feasible if not infeasible, even for moderate samples sizes. Precisely, it involves averaging O(n d1+...+dK ) terms, when considering a U-statistic of degrees (d1, . . . , dK) based on samples of sizes proportional to n. We propose here to consider a drastically simpler Monte-Carlo version of the empirical risk based on O(n) terms solely, which can be viewed as an incomplete generalized U-statistic, and prove that, remarkably, the approximation stage does not damage the ERM procedure and yields a learning rate of order OP(1/ √ n). Beyond a theoretical analysis guaranteeing the validity of this approach, numerical experiments are displayed for illustrative purpose.

Dates et versions

hal-01242334 , version 1 (11-12-2015)

Identifiants

Citer

Stéphan Clémençon, Sylvain Robbiano, Jessica Tressou. Maximal deviations of incomplete U-statistics with applications to empirical risk sampling. SIAM International Conference on Data-Mining SDM13, May 2013, Austin, United States. pp.19-27, ⟨10.1137/1.9781611972832.3⟩. ⟨hal-01242334⟩
69 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More