ERM and RERM are optimal estimators for regression problems when malicious outliers corrupt the labels - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Article Dans Une Revue Electronic Journal of Statistics Année : 2020

ERM and RERM are optimal estimators for regression problems when malicious outliers corrupt the labels

Résumé

We study Empirical Risk Minimizers (ERM) and Regularized Empirical Risk Minimizers (RERM) for regression problems with convex and L-Lipschitz loss functions. We consider a setting where |O| malicious outliers contaminate the labels. In that case, under a local Bernstein condition, we show that the L 2-error rate is bounded by r N + AL|O|/N , where N is the total number of observations, r N is the L 2-error rate in the noncontaminated setting and A is a parameter coming from the local Bernstein condition. When r N is minimax-rate-optimal in a non-contaminated setting, the rate r N +AL|O|/N is also minimax-rate-optimal when |O| outliers contaminate the label. The main results of the paper can be used for many non-regularized and regularized procedures under weak assumptions on the noise. We present results for Huber's M-estimators (without penalization or regularized by the 1-norm) and for general regularized learning problems in reproducible kernel Hilbert spaces when the noise can be heavy-tailed.
Fichier principal
Vignette du fichier
20-EJS1754 (8).pdf (826.32 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-03874524 , version 1 (28-11-2022)

Identifiants

Citer

Geoffrey Chinot. ERM and RERM are optimal estimators for regression problems when malicious outliers corrupt the labels. Electronic Journal of Statistics , 2020, 14, ⟨10.1214/20-ejs1754⟩. ⟨hal-03874524⟩
12 Consultations
19 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More