Contextual Bandit for Active Learning: Active Thompson Sampling

Djallel Bouneffouf 1, 2 Romain Laroche 3 Tanguy Urvoy 3 Raphael Féraud 3 Robin Allesiardo 4, 3
4 TAO - Machine Learning and Optimisation
LRI - Laboratoire de Recherche en Informatique, UP11 - Université Paris-Sud - Paris 11, Inria Saclay - Ile de France, CNRS - Centre National de la Recherche Scientifique : UMR8623
Abstract : The labelling of training examples is a costly task in a supervised classi cation. Active learning strategies answer this problem by selecting the most useful unlabelled examples to train a predictive model. The choice of examples to label can be seen as a dilemma between the exploration and the exploitation over the data space representation. In this paper, a novel active learning strategy manages this compromise by modelling the active learning problem as a contextual bandit problem. We propose a sequential algorithm named Active Thompson Sampling (ATS), which, in each round, assigns a sampling distribution on the pool, samples one point from this distribution, and queries the oracle for this sample point label. Experimental comparison to previously proposed active learning algorithms show superior performance on a real application dataset.
Type de document :
Pré-publication, Document de travail
21st International Conference on Neural Information Processing. 2014
Liste complète des métadonnées

Littérature citée [11 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01069802
Contributeur : Djallel Bouneffouf <>
Soumis le : lundi 29 septembre 2014 - 19:58:39
Dernière modification le : jeudi 5 avril 2018 - 12:30:12
Document(s) archivé(s) le : mardi 30 décembre 2014 - 11:50:29

Fichier

Contextual_Bandit_for_Active_L...
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01069802, version 1

Collections

Citation

Djallel Bouneffouf, Romain Laroche, Tanguy Urvoy, Raphael Féraud, Robin Allesiardo. Contextual Bandit for Active Learning: Active Thompson Sampling. 21st International Conference on Neural Information Processing. 2014. 〈hal-01069802〉

Partager

Métriques

Consultations de la notice

2289

Téléchargements de fichiers

3892