Consiglio Nazionale delle Ricerche

Tipo di prodottoArticolo in rivista
TitoloHow to learn multiple tasks
Anno di pubblicazione2008
FormatoCartaceo
Autore/iCalabretta, R., Di Ferdinando, A., Parisi, D., Keil, F. C.
Affiliazioni autoriISTC-Cnr ISTC-CnrISTC-Cnr Yale University
Autori CNR e affiliazioni
  • DOMENICO PARISI
  • RAFFAELE CALABRETTA
Lingua/e
  • inglese
AbstractThis paper examines the question of how the presence of multiple tasks interacts with learning architectures and the flow of information through those architectures. It approaches the question by using the idealization of an artificial neural network where it is possible to ask more precisely about the effects of modular versus non-modular architectures as well as the effects of sequential vs. simultaneous learning of tasks. While prior work has shown a clear advantage of modular architectures when the two tasks must be learned at the same time from the start, this advantage may disappear when one task is first learned to a criterion before the second task is undertaken. Nonmodular networks, in some cases of sequential learning, achieve success levels comparable to those of modular networks. In particular, if a nonmodular network is to learn two tasks of different difficulties and the more difficult task is presented first and learned to a criterion, then the network will learn the second easier one without permanent degradation of the first one. In contrast, if the easier task is learned first, a nonmodular task may perform significantly less well than a modular one. It seems that the reason for these difference has to do with the fact that the sequential presentation of the more difficult task first minimizes interference between the two tasks. More broadly, the studies summarized in this paper make it clear no one learning architecture is optimal in all situations.
Lingua abstract-
Altro abstract-
Lingua altro abstract-
Pagine da30
Pagine a41
Pagine totali-
RivistaBiological theory
Attiva dal 2006
Editore: MIT Press, - Cambridge, MA
Paese di pubblicazione: Stati Uniti d'America
Lingua: inglese
ISSN: 1555-5542
Titolo chiave: Biological theory
Titolo proprio: Biological theory.
Titolo abbreviato: Biol. theory
Numero volume della rivista3
Fascicolo della rivista-
DOI-
Verificato da referee-
Stato della pubblicazione-
Indicizzazione (in banche dati controllate)-
Parole chiavegenetic interference, evolution of modularity, what and where task, evolutionary connectionism, evolvability
Link (URL, URI)-
Titolo parallelo-
Licenza-
Scadenza embargo-
Data di accettazione-
Note/Altre informazioni-
Strutture CNR
  • ISTC — Istituto di scienze e tecnologie della cognizione
Moduli/Attività/Sottoprogetti CNR
  • ICT.P08.006.001 : Tecnologie avanzate per l'interazione uomo, robot ed agenti intelligenti
Progetti Europei-
Allegati

Dati storici
I dati storici non sono modificabili, sono stati ereditati da altri sistemi (es. Gestione Istituti, PUMA, ...) e hanno solo valore storico.
Area disciplinarePsychology
Area valutazione CIVRScienze storiche, filosofiche, psicologiche e pedagogiche
Rivista ISIBIOLOGICAL THEORY [00122NN]
NoteBiological Theory (MIT Press)