Importance Reweighting for Biquality Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Importance Reweighting for Biquality Learning

Pierre Nodet
Vincent Lemaire
  • Fonction : Auteur
Alexis Bondu
  • Fonction : Auteur
Adam Ouorou
  • Fonction : Auteur

Résumé

The field of Weakly Supervised Learning (WSL) has recently seen a surge of popularity, with numerous papers addressing different types of "supervision deficiencies", namely: poor quality, non adaptability, and insufficient quantity of labels. Regarding quality, label noise can be of different types, including completely-at-random, at-random or even not-at-random. All these kinds of label noise are addressed separately in the literature, leading to highly specialized approaches. This paper proposes an original, encompassing, view of Weakly Supervised Learning, which results in the design of generic approaches capable of dealing with any kind of label noise. For this purpose, an alternative setting called "Biquality data" is used. It assumes that a small trusted dataset of correctly labeled examples is available, in addition to an untrusted dataset of noisy examples. In this paper, we propose a new reweigthing scheme capable of identifying noncorrupted examples in the untrusted dataset. This allows one to learn classifiers using both datasets. Extensive experiments that simulate several types of label noise and that vary the quality and quantity of untrusted examples, demonstrate that the proposed approach outperforms baselines and state-ofthe-art approaches.
Fichier principal
Vignette du fichier
2010.09621.pdf (536.85 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03650243 , version 1 (24-04-2022)

Identifiants

Citer

Pierre Nodet, Vincent Lemaire, Alexis Bondu, Antoine Cornuejols, Adam Ouorou. Importance Reweighting for Biquality Learning. 2021 International Joint Conference on Neural Networks (IJCNN), Jul 2021, Shenzhen, China. pp.1-8, ⟨10.1109/IJCNN52387.2021.9533349⟩. ⟨hal-03650243⟩
32 Consultations
31 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More