Communication Dans Un Congrès Année : 2025

Model selection for behavioral learning data and applications to contextual bandits

Julien Aubert
Louis Köhler
  • Fonction : Auteur
  • PersonId : 1105860
Giulia Mezzadri
  • Fonction : Auteur
  • PersonId : 1207986
Patricia Reynaud-Bouret

Résumé

Learning for animals or humans is the process that leads to behaviors better adapted to the environment. This process highly depends on the individual that learns and is usually observed only through the individual's actions. This article presents ways to use this individual behavioral data to find the model that best explains how the individual learns. We propose two model selection methods: a general hold-out procedure and an AIC-type criterion, both adapted to non-stationary dependent data. We provide theoretical error bounds for these methods that are close to those of the standard i.i.d. case. To compare these approaches, we apply them to contextual bandit models and illustrate their use on both synthetic and experimental learning data in a human categorization task.
Fichier principal
Vignette du fichier
aistats_2025.pdf (716) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04946461 , version 1 (18-02-2025)

Licence

Identifiants

  • HAL Id : hal-04946461 , version 1

Citer

Julien Aubert, Louis Köhler, Luc Lehéricy, Giulia Mezzadri, Patricia Reynaud-Bouret. Model selection for behavioral learning data and applications to contextual bandits. 28th International Conference on Artificial Intelligence and Statistics (AISTATS), May 2025, Mai Khao, Thailand. ⟨hal-04946461⟩
0 Consultations
0 Téléchargements

Partager

More