<div class="csl-bib-body">
<div class="csl-entry">Pasteris, S., Rumi, A., Thiessen, M., Saito, S., Miyauchi, A., Vitale, F., & Herbster, M. (2024, June 17). <i>Bandits with Abstention under Expert Advice</i> [Poster Presentation]. ICML 2024 Workshop: Foundations of Reinforcement Learning and Control -- Connections and Perspectivesing (ICML 2024), Vienna, Austria. http://hdl.handle.net/20.500.12708/199835</div>
</div>
-
dc.identifier.uri
http://hdl.handle.net/20.500.12708/199835
-
dc.description.abstract
We study the classic problem of prediction with expert advice under bandit feedback. Our model assumes that one action, corresponding to the learner’s abstention from play, has no reward or loss on every trial. We propose the confidence-rated bandits with abstentions (CBA) algorithm, which exploits this assumption to obtain reward bounds that can significantly improve those of the classical EXP4 algorithm. Our problem can be construed as the aggregation of confidence-rated predictors, with the learner having the option to abstain from play. We are the first to achieve bounds on the expected cumulative reward for general confidence-rated predictors. In the special case of specialists we achieve a novel reward bound, significantly improving previous bounds of SPECIALISTEXP (treating abstention as another action). We discuss how CBA can be applied to the problem of adversarial contextual bandits with the option of abstaining from selecting any action. We are able to leverage a wide range of inductive biases, outperforming previous approaches both theoretically and in preliminary experimental analysis. Additionally, we achieve a reduction in runtime from quadratic to almost linear in the number of contexts for the specific case of metric space contexts.
en
dc.language.iso
en
-
dc.subject
Machine Learning
en
dc.subject
Bandits
en
dc.title
Bandits with Abstention under Expert Advice
en
dc.type
Presentation
en
dc.type
Präsentation
de
dc.contributor.affiliation
Turing Institute, United Kingdom of Great Britain and Northern Ireland (the)
-
dc.contributor.affiliation
University of Milan, Italy
-
dc.contributor.affiliation
University College London, United Kingdom of Great Britain and Northern Ireland (the)
-
dc.contributor.affiliation
CENTAI Institute, Italy
-
dc.contributor.affiliation
CENTAI Institute, Italy
-
dc.contributor.affiliation
University College London, United Kingdom of Great Britain and Northern Ireland (the)
-
dc.type.category
Poster Presentation
-
tuw.researchTopic.id
I4
-
tuw.researchTopic.name
Information Systems Engineering
-
tuw.researchTopic.value
100
-
tuw.linking
https://openreview.net/forum?id=2h3YuVD6NS
-
tuw.publication.orgunit
E194-06 - Forschungsbereich Machine Learning
-
tuw.author.orcid
0000-0001-9333-2685
-
tuw.author.orcid
0000-0003-0722-6195
-
tuw.author.orcid
0000-0002-4673-7476
-
tuw.event.name
ICML 2024 Workshop: Foundations of Reinforcement Learning and Control -- Connections and Perspectivesing (ICML 2024)
en
tuw.event.startdate
26-07-2024
-
tuw.event.enddate
26-07-2024
-
tuw.event.online
On Site
-
tuw.event.type
Event for scientific audience
-
tuw.event.place
Vienna
-
tuw.event.country
AT
-
tuw.event.presenter
Thiessen, Maximilian
-
wb.sciencebranch
Informatik
-
wb.sciencebranch.oefos
1020
-
wb.sciencebranch.value
100
-
item.openairecristype
http://purl.org/coar/resource_type/c_18co
-
item.openairetype
conference poster not in proceedings
-
item.fulltext
no Fulltext
-
item.languageiso639-1
en
-
item.grantfulltext
restricted
-
item.cerifentitytype
Publications
-
crisitem.author.dept
Turing Institute
-
crisitem.author.dept
University of Milan
-
crisitem.author.dept
E194-06 - Forschungsbereich Machine Learning
-
crisitem.author.dept
University College London
-
crisitem.author.dept
CENTAI Institute, Italy
-
crisitem.author.dept
CENTAI Institute, Italy
-
crisitem.author.dept
University College London
-
crisitem.author.orcid
0000-0001-9333-2685
-
crisitem.author.orcid
0000-0003-0722-6195
-
crisitem.author.orcid
0000-0002-4673-7476
-
crisitem.author.parentorg
E194 - Institut für Information Systems Engineering
