Maturationally-Constrained Competence-Based Intrinsically Motivated Learning

Adrien Baranes 1, * Pierre-Yves Oudeyer 1
* Auteur correspondant
1 Flowers - Flowing Epigenetic Robots and Systems
Inria Bordeaux - Sud-Ouest, U2IS - Unité d'Informatique et d'Ingénierie des Systèmes
Abstract : This paper studies the coupling of intrinsic motiva- tion and physiological maturational constraints, and argues that both mechanisms may have complex bidirectional interactions allowing to actively control the growth of complexity in motor development. First, we introduce the self-adaptive goal generation algorithm (SAGG), instantiating an intrinsically motivated goal exploration mechanism for motor learning of inverse models. Then, we introduce a functional model of maturational con- straints inspired by the myelination process in humans, and show how it can be coupled with the SAGG algorithm, forming a new system called McSAGG. We then present experiments to evaluate qualitative properties of these systems when applied to learning a reaching skill with an arm with initially unknown kinematics.
Type de document :
Communication dans un congrès
IEEE International Conference on Development and Learning (ICDL 2010), 2010, Ann Arbor, United States. 2010
Liste complète des métadonnées

Littérature citée [26 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/inria-00541770
Contributeur : Pierre Rouanet <>
Soumis le : mercredi 1 décembre 2010 - 10:58:12
Dernière modification le : jeudi 16 novembre 2017 - 17:12:01
Document(s) archivé(s) le : mercredi 2 mars 2011 - 02:58:00

Fichier

BaranesOudeyerICDL10.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : inria-00541770, version 1

Collections

Citation

Adrien Baranes, Pierre-Yves Oudeyer. Maturationally-Constrained Competence-Based Intrinsically Motivated Learning. IEEE International Conference on Development and Learning (ICDL 2010), 2010, Ann Arbor, United States. 2010. 〈inria-00541770〉

Partager

Métriques

Consultations de la notice

254

Téléchargements de fichiers

128