Scalable reinforcement learning through hierarchical decompositions for weakly-coupled problems
DC Element | Wert | Sprache |
---|---|---|
dc.contributor.author | Toutounji, H. | |
dc.contributor.author | Rothkopf, C.A. | |
dc.contributor.author | Triesch, J. | |
dc.date.accessioned | 2021-12-23T16:31:17Z | - |
dc.date.available | 2021-12-23T16:31:17Z | - |
dc.date.issued | 2011 | |
dc.identifier.isbn | 9781612849904 | |
dc.identifier.uri | https://osnascholar.ub.uni-osnabrueck.de/handle/unios/16998 | - |
dc.description | Conference of 2011 IEEE International Conference on Development and Learning, ICDL 2011 ; Conference Date: 24 August 2011 Through 27 August 2011; Conference Code:87020 | |
dc.description.abstract | Reinforcement Learning, or Reward-Dependent Learning, has been very successful at describing how animals and humans adjust their actions so as to increase their gains and reduce their losses in a wide variety of tasks. Empirical studies have furthermore identified numerous neuronal correlates of quantities necessary for such computations. But, in general it is too expensive for the brain to encode actions and their outcomes with respect to all available dimensions describing the state of the world. This suggests the existence of learning algorithms that are capable of taking advantage of the independencies present in the world and hence reducing the computational costs in terms of representations and learning. A possible solution is to use separate learners for task dimensions with independent dynamics and rewards. But the condition of independence is usually too restrictive. Here, we propose a hierarchical reinforcement learning solution for the more general case in which the dynamics are not independent but weakly coupled and show how to assign credit to the different modules, which solve the task jointly. © 2011 IEEE. | |
dc.description.sponsorship | Seventh Framework ProgrammeSeventh Framework Programme,FP7,231722; IEEE Computational Intelligence Society; Frankfurt Institute for Advanced Studies (FIAS); Bielefeld Univ., Cognitive Interact.; Technol. Cent. Excellence (CITEC); italk | |
dc.language.iso | en | |
dc.relation.ispartof | 2011 IEEE International Conference on Development and Learning, ICDL 2011 | |
dc.subject | Computational costs | |
dc.subject | Empirical studies | |
dc.subject | Hierarchical decompositions | |
dc.subject | Hierarchical reinforcement learning | |
dc.subject | Possible solutions | |
dc.subject | Task dimensions, Animals | |
dc.subject | Learning algorithms, Reinforcement learning | |
dc.title | Scalable reinforcement learning through hierarchical decompositions for weakly-coupled problems | |
dc.type | conference paper | |
dc.identifier.doi | 10.1109/DEVLRN.2011.6037351 | |
dc.identifier.scopus | 2-s2.0-80055002863 | |
dc.identifier.url | https://www.scopus.com/inward/record.uri?eid=2-s2.0-80055002863&doi=10.1109%2fDEVLRN.2011.6037351&partnerID=40&md5=167003b600ed2ae77ace6f98bc3d7356 | |
dc.publisher.place | Frankfurt am Main | |
dcterms.isPartOf.abbreviation | IEEE Int. Conf. Dev. Learn., ICDL |
Seitenaufrufe
1
Letzte Woche
0
0
Letzter Monat
0
0
geprüft am 07.06.2024