Show simple item record

dc.contributor.authorCliment Muñoz, Antoni
dc.contributor.authorGnatyshak, Dmitry
dc.contributor.authorÁlvarez Napagao, Sergio
dc.contributor.otherUniversitat Politècnica de Catalunya. Doctorat en Intel·ligència Artificial
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament de Ciències de la Computació
dc.contributor.otherBarcelona Supercomputing Center
dc.date.accessioned2021-11-16T13:30:48Z
dc.date.available2021-11-16T13:30:48Z
dc.date.issued2021
dc.identifier.citationCliment, A.; Gnatyshak, D.; Álvarez-Napagao, S. Applying and verifying an explainability method based on policy graphs in the context of reinforcement learning. A: International Conference of the Catalan Association for Artificial Intelligence. "Artificial Intelligence Research and Development: proceedings of the 23rd International Conference of the Catalan Association for Artificial Intelligence". IOS Press, 2021, p. 455-464. ISBN 978-1-64368-211-2. DOI 10.3233/FAIA210166.
dc.identifier.isbn978-1-64368-211-2
dc.identifier.urihttp://hdl.handle.net/2117/356542
dc.description.abstractThe advancement on explainability techniques is quite relevant in the field of Reinforcement Learning (RL) and its applications can be beneficial for the development of intelligent agents that are understandable by humans and are able cooperate with them. When dealing with Deep RL some approaches already exist in the literature, but a common problem is that it can be tricky to define whether the explanations generated for an agent really reflect the behaviour of the trained agent. In this work we will apply an approach for explainability based on the creation of a Policy Graph (PG) that represents the agent’s behaviour. Our main contribution is a way to measure the similarity between the explanations and the agent’s behaviour, by building another agent that follows a policy based on the explainability method and comparing the behaviour of both agents.
dc.format.extent10 p.
dc.language.isoeng
dc.publisherIOS Press
dc.rightsAttribution-NonCommercial 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by-nc/4.0/
dc.subjectÀrees temàtiques de la UPC::Informàtica::Intel·ligència artificial::Agents intel·ligents
dc.subject.lcshIntelligent agents (Computer software)
dc.subject.lcshReinforcement learning
dc.subject.otherExplainable AI
dc.subject.otherPolicy graphs
dc.titleApplying and verifying an explainability method based on policy graphs in the context of reinforcement learning
dc.typeConference report
dc.subject.lemacAgents intel·ligents (Programari)
dc.subject.lemacAprenentatge per reforç
dc.contributor.groupUniversitat Politècnica de Catalunya. KEMLG - Grup d'Enginyeria del Coneixement i Aprenentatge Automàtic
dc.identifier.doi10.3233/FAIA210166
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttps://ebooks.iospress.nl/doi/10.3233/FAIA210166
dc.rights.accessOpen Access
local.identifier.drac32211581
dc.description.versionPostprint (published version)
local.citation.authorCliment, A.; Gnatyshak, D.; Álvarez-Napagao, S.
local.citation.contributorInternational Conference of the Catalan Association for Artificial Intelligence
local.citation.publicationNameArtificial Intelligence Research and Development: proceedings of the 23rd International Conference of the Catalan Association for Artificial Intelligence
local.citation.startingPage455
local.citation.endingPage464


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial 4.0 Generic
Except where otherwise noted, content on this work is licensed under a Creative Commons license : Attribution-NonCommercial 4.0 Generic