BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Date iCal//NONSGML kigkonsult.se iCalcreator 2.20.2//
METHOD:PUBLISH
X-WR-CALNAME;VALUE=TEXT:Eventi DIAG
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:STANDARD
DTSTART:20221030T030000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20230326T020000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:calendar.25622.field_data.0@www.ugov-ricerca.uniroma1.it
DTSTAMP:20260410T022836Z
CREATED:20230213T080618Z
DESCRIPTION:Link Zoom: https://uniroma1.zoom.us/j/93674593746?pwd=N2xqNzJWe
 HA0b2V1dUlxYktlMUpkdz09 Abstract: In adversarial team games\, a team of pl
 ayers sequentially faces a team of adversaries. These games are the simple
 st setting with multiple players where cooperation and competition coexist
 \, and it is known that the information asymmetry among the team members m
 akes equilibrium approximation computationally hard. Although much effort 
 has been spent designing scalable algorithms\, the problem of solving larg
 e game instances is open. This work shows that we can recover from this we
 akness by bridging the gap between sequential adversarial team games and 2
 -player games. In particular\, we propose a new\, suitable game representa
 tion that we call team public information\, in which a team is represented
  as a single coordinator who only knows information common to the whole te
 am and prescribes to each member an action for any possible private state.
  The resulting representation is highly explainable\, being a 2-player tre
 e in which the team’s strategies are behavioral with a direct interpretati
 on and more expressive than the original extensive form when designing abs
 tractions. Furthermore\, we prove the payoff equivalence of our representa
 tion\, and we provide techniques that\, starting directly from the extensi
 ve form\, generate dramatically more compact representations without infor
 mation loss. Finally\, we experimentally evaluate our techniques when appl
 ied to a standard testbed\, comparing their performance with the current s
 tate of the art.  Bio sketch: Nicola Gatti is an associate professor of Co
 mputer Science and Engineering in the Department of Electronics\, Informat
 ion\, and Bioengineering at Politecnico di Milano. His main achievements c
 ome from algorithmic game theory\, allocation problems and incentives\, al
 gorithmic social choice theory\, multiagent learning\, and online learning
 . He received several awards\, including the 2011 AIxIA Marco Somalvico Aw
 ard as the best Italian young researcher in AI\, the best paper award in s
 everal conferences\, including the prestigious NeurIPS 2020 and Cooperativ
 e AI 2021 funded by Google Deepmind. In 2021 he was elected as a EurAi Fel
 low (top
DTSTART;TZID=Europe/Paris:20230217T150000
DTEND;TZID=Europe/Paris:20230217T150000
LAST-MODIFIED:20230215T063343Z
LOCATION:DIAG - Aula Magna
SUMMARY:Recent Advancements in Equilibrium Computation for Adversarial Team
  Games - Nicola Gatti (Politecnico di Milano)
URL;TYPE=URI:http://www.ugov-ricerca.uniroma1.it/node/25622
END:VEVENT
END:VCALENDAR