Deep Multi-Agent Reinforcement Learning in a Homogeneous Open Population

Roxana Radulescu, Manon Legrand, Kyriakos Efthymiadis, Diederik Roijers, Ann Nowe

Research output: Chapter in Book/Report/Conference proceedingConference paper

7 Citations (Scopus)
235 Downloads (Pure)

Abstract

Advances in reinforcement learning research have recently produced agents that are competent, or sometimes exceed human performance, in complex tasks. Most interesting real world problems however, are not restricted to one agent, but instead deal with multiple agents acting in the same environment and have proven to be challenging tasks to solve. In this work we present a study on a homogeneous open population of agents modelled as a multi-agent reinforcement learning (MARL) system. We propose a centralised learning approach, with decentralised execution in which agents are given the same policy to execute individually. Using the SimuLane highway traffic simulator as a test-bed we show experimentally that using a single-agent learnt policy to initialise the multi-agent scenario, which we then fine-tune to the task, out-performs agents that learn in the multi-agent setting from scratch. Specifically we contribute an open population MARL configuration, how to transfer knowledge from single- to a multi-agent setting and a training procedure for a homogeneous open population of agents.

Original languageEnglish
Title of host publicationArtificial Intelligence
Subtitle of host publication30th Benelux Conference, BNAIC 2018, ‘s-Hertogenbosch, The Netherlands, November 8–9, 2018, Revised Selected Papers
EditorsMartin Atzmueller, Wouter Duivesteijn
PublisherSpringer International Publishing
Pages177-191
Number of pages15
ISBN (Electronic)978-3-030-31978-6
ISBN (Print)978-3-030-31977-9
DOIs
Publication statusPublished - 8 Nov 2018
Event30th Benelux Conference on Artificial Intelligence - ‘s-Hertogenbosch, Netherlands
Duration: 8 Nov 20189 Nov 2018
https://bnaic2018.nl

Publication series

NameBelgian/Netherlands Artificial Intelligence Conference
ISSN (Print)1568-7805

Conference

Conference30th Benelux Conference on Artificial Intelligence
Abbreviated titleBNAIC 2018
Country/TerritoryNetherlands
City‘s-Hertogenbosch
Period8/11/189/11/18
Internet address

Fingerprint

Dive into the research topics of 'Deep Multi-Agent Reinforcement Learning in a Homogeneous Open Population'. Together they form a unique fingerprint.

Cite this