Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-8423
Full metadata record
DC FieldValueLanguage
dc.contributor.authorSezener, Can Eren-
dc.contributor.authorDezfouli, Amir-
dc.contributor.authorKeramati, Mehdi-
dc.date.accessioned2019-04-29T11:28:10Z-
dc.date.available2019-04-29T11:28:10Z-
dc.date.issued2019-03-12-
dc.identifier.issn1553-734X-
dc.identifier.urihttps://depositonce.tu-berlin.de/handle/11303/9365-
dc.identifier.urihttp://dx.doi.org/10.14279/depositonce-8423-
dc.description.abstractEvaluating the future consequences of actions is achievable by simulating a mental search tree into the future. Expanding deep trees, however, is computationally taxing. Therefore, machines and humans use a plan-until-habit scheme that simulates the environment up to a limited depth and then exploits habitual values as proxies for consequences that may arise in the future. Two outstanding questions in this scheme are “in which directions the search tree should be expanded?”, and “when should the expansion stop?”. Here we propose a principled solution to these questions based on a speed/accuracy tradeoff: deeper expansion in the appropriate directions leads to more accurate planning, but at the cost of slower decision-making. Our simulation results show how this algorithm expands the search tree effectively and efficiently in a grid-world environment. We further show that our algorithm can explain several behavioral patterns in animals and humans, namely the effect of time-pressure on the depth of planning, the effect of reward magnitudes on the direction of planning, and the gradual shift from goal-directed to habitual behavior over the course of training. The algorithm also provides several predictions testable in animal/human experiments.en
dc.description.sponsorshipDFG, 414044773, Open Access Publizieren 2019 - 2020 / Technische Universität Berlinen
dc.language.isoenen
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en
dc.subject.ddc570 Biowissenschaften; Biologiede
dc.subject.ddc004 Datenverarbeitung; Informatikde
dc.subject.ddc610 Medizin und Gesundheitde
dc.subject.othersearch treeen
dc.subject.otherMonte Carlo methoden
dc.subject.otherdecision makingen
dc.subject.otherbehavioral patternen
dc.titleOptimizing the depth and the direction of prospective planning using information valuesen
dc.typeArticleen
tub.accessrights.dnbfreeen
tub.publisher.universityorinstitutionTechnische Universität Berlinen
dc.identifier.eissn1553-7358-
dc.type.versionpublishedVersionen
dcterms.bibliographicCitation.doi10.1371/journal.pcbi.1006827en
dcterms.bibliographicCitation.journaltitlePLOS Computational Biologyen
dcterms.bibliographicCitation.originalpublisherplaceSan Francisco, Calif.en
dcterms.bibliographicCitation.volume15en
dcterms.bibliographicCitation.originalpublishernamePublic Library of Scienceen
dcterms.bibliographicCitation.issue3en
dcterms.bibliographicCitation.articlenumbere1006827en
Appears in Collections:FG Neuronale Informationsverarbeitung » Publications

Files in This Item:
File Description SizeFormat 
journal.pcbi.1006827.pdf1.72 MBAdobe PDFView/Open


This item is licensed under a Creative Commons License Creative Commons