Evolving internal reinforcers for an intrinsically motivated reinforcement-learning robot (Contributo in volume (capitolo o saggio))

Type
Label
  • Evolving internal reinforcers for an intrinsically motivated reinforcement-learning robot (Contributo in volume (capitolo o saggio)) (literal)
Anno
  • 2007-01-01T00:00:00+01:00 (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#doi
  • 10.1109/DEVLRN.2007.4354052 (literal)
Alternative label
  • Schembri M., Mirolli M., Baldassarre G. (2007)
    Evolving internal reinforcers for an intrinsically motivated reinforcement-learning robot
    Imperial College, London (Regno Unito) in IEEE 6th International Conference on Development and Learning (ICDL2007), 2007
    (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#autori
  • Schembri M., Mirolli M., Baldassarre G. (literal)
Pagina inizio
  • 282 (literal)
Pagina fine
  • 287 (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#citta
  • London (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#url
  • http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4354052 (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#titoloVolume
  • IEEE 6th International Conference on Development and Learning (ICDL2007) (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#note
  • in Demiris Y., Scassellati B., Mareschal D. (eds.), E1-6.,pp. 282-287. (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#pagineTotali
  • 6 (literal)
Note
  • ISI Web of Science (WOS) (literal)
  • Google Scholar (literal)
  • Scopu (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#affiliazioni
  • Istituto di scienze e tecnologie della cognizione (literal)
Titolo
  • Evolving internal reinforcers for an intrinsically motivated reinforcement-learning robot (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#inCollana
  • The 6th IEEE International Conference on Development and Learning (ICDL2007) (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#isbn
  • 978-1-4244-1116-0 (literal)
Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#curatoriVolume
  • Demiris Y.; Scassellati B.; Mareschal D. (literal)
Abstract
  • Intrinsically Motivated Reinforcement Learning (IMRL) has been proposed as a framework within which agents exploit \"internal reinforcement\" to acquire general-purpose building-block behaviors (\"skills\") which can be later combined for solving several specific tasks. The architectures so far proposed within this framework are limited in that: (1) they use hardwired \"salient events\" to form and train skills, and this limits agents' autonomy; (2) they are applicable only to problems with abstract states and actions, as grid-world problems. This paper proposes solutions to these problems in the form of a hierarchical reinforcement-learning architecture that: (1) exploits the ideas and techniques of Evolutionary Robotics to allow the system to autonomously discover \"salient events\"; (2) uses neural networks to allow the system to cope with continuous states and noisy environments. The paper also starts to explore a new way of producing intrinsic motivations on the basis of the learning progress of skills. The viability of the proposed approach is demonstrated with a simulated robotic scenario. (literal)
Editore
Prodotto di
Autore CNR

Incoming links:


Prodotto
Autore CNR di
Editore di
data.CNR.it