BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260901T200341EDT-50042fdDbk@132.216.98.100
DTSTAMP:20260902T000341Z
DESCRIPTION:Abstract\n\nExploring the process by which autonomous agents co
 ordinate represents a pivotal advancement toward emulating populations\, w
 hich encompasses diverse applications in robotics\, game theory\, economic
 s\, and social sciences. Specifically\, the mechanisms that allow cooperat
 ing learners to converge toward coherent team strategies remain poorly und
 erstood\, and yet\, understudied. The myriad of challenges that hinder con
 current learning -- environment non-stationarity\, intricate credit assign
 ments\, exponential complexity\, etc. -- has fragmented the focus of the l
 iterature such that the underlying mechanisms of successful multi-agent co
 ordination remain to be pinpointed. This work puts forward and explores th
 ree key elements of coordination\, namely\, shared incentives\, interactio
 ns\, and the use of internal models. We scrutinize this take by inspecting
  how these components can enable coordination in several multi-agent learn
 ing paradigms that cover the well-established Multi-Agent Reinforcement Le
 arning framework\, its offline interaction-less variation\, and the Archit
 ect-Builder Problem\, a novel reward-less interactive learning setting. Ou
 r research derives fresh insights into fostering coordination and while th
 e implications of multi-agent learning extend across various fields\, we a
 re particularly interested in planning solutions to societal and ecologica
 l challenges by simulating how populations would react to changes in their
  environment.\n
DTSTART:20231211T150000Z
DTEND:20231211T170000Z
LOCATION:Room 603\, McConnell Engineering Building\, CA\, QC\, Montreal\, H
 3A 0E9\, 3480 rue University
SUMMARY:PhD defence of Paul Barde – Coordination in Cooperative Multi-Agent
  Learning
URL:https://www.mcgill.ca/ece/channels/event/phd-defence-paul-barde-coordin
 ation-cooperative-multi-agent-learning-353367
END:VEVENT
END:VCALENDAR