BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Ad-hoc Coordination in Anonymous Games
DTSTART:20170531T140000
DTEND:20170531T160000
DTSTAMP:20260407T070736Z
UID:fb3658d0f2faeab6105e59a506020c20f7c428d9369af796cf90c38f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Panayiotis Danassis\nEDIC Candidacy Exam\nExam president: Prof
 . Alcherio Martinoli\nThesis advisor: Prof. Boi Faltings\nCo-examiner: Pro
 f. Patrick Thiran\n\nAbstract\nAd-hoc multi-agent coordination is a relati
 vely new research area\, recently introduced in the seminal work of Stone 
 et al. [Stone 2010]. Recent advances in domains such as Autonomous Agents\
 , Intelligent Robotics\, Internet of Things and Cyber-Physical Systems res
 ulted in an immense growth of autonomous software and robotic agents. As a
 utonomous agents continue to proliferate\, so does the need for them to in
 teract and coordinate efficiently. However\, due to the differences betwee
 n the agents in terms of origin\, reasoning\, knowledge\, and perceptual a
 nd actuation capabilities\, such teamwork must take place without an a pri
 ori defined coordination protocol or perhaps any form of explicit communic
 ation. This is in contrast to most of the prior research done on multi-age
 nt teamwork\, which requires explicit coordination protocols\, and/or shar
 ed assumptions. This emphasizes the need to develop novel and robust intel
 ligent agents that are able to engage in ad-hoc teamwork and coordination 
 and efficiently cooperate with previously unencountered agents.\n\nThis th
 esis aims to investigate the problem of ad-hoc coordination in anonymous g
 ames. In algorithmic game theory\, anonymous games are multi-agent games i
 n which an agent does not distinguish between other agents\, meaning that 
 his utility depends on his own strategy as well as the number of agents th
 at choose each of the other strategies (and not to the identities of the o
 ther agents). We are interested in such games because of their ability to 
 capture a wide range of real world phenomena\, having applications in reso
 urce sharing\, foraging settings\, market settings etc. Ad-hoc problems th
 ough suffer from high dimensionality [Barrett 2012]. Hence\, contrary to c
 onventional approaches in the domain of ad-hoc coordination which usually 
 involve solving large Partially Observable Markov Decision Processes (POMD
 Ps) or arduous Bayesian Learning\, our goal is to develop robust and simpl
 e dynamics. Such an example is the decentralized algorithm of Cigler et al
 . [Cigler 2011] where the agents are able to reach an efficient and fair c
 orrelated equilibrium through repeated interactions and multi-agent learni
 ng.\n\nBackground papers\nAd hoc autonomous agent teams: Collaboration wit
 hout pre-coordination by P. Stone\, G. A. Kaminka\, S. Kraus\, and J. S. R
 osenschein.\nAn analysis framework for ad hoc teamwork tasks by S. Barrett
  and P. Stone.\nReaching correlated equilibria through multi-agent learnin
 g by L. Cigler and B. Faltings\,
LOCATION:INR 212 https://plan.epfl.ch/theme/generalite_thm_plan_public?lan
 g=en&room=inr%20212&dim_floor=2&dim_lang=en&baselayer_ref=grp_backgrounds&
 tree_groups=centres_nevralgiques%2Cacces%2Cmobilite_reduite%2Censeign
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR