Analysing Congestion Problems in Multi-agent Reinforcement Learning

Onderzoeksoutput: Conference paper

7 Citaten (Scopus)
7 Downloads (Pure)

Samenvatting

We extend the study of congestion problems to a more realistic scenario, the Road Network Domain (RND), where the resources are no longer independent, but rather part of a network, thus choosing one path will also impact the load of another one having common road segments. We demonstrate the application of state-of-the-art multi-agent reinforcement learning methods for this new congestion model and analyse their performance. RND allows us to highlight an important limitation of resource abstraction and show that the difference rewards approach manages to better capture and inform the agents about the dynamics of the environment.

Originele taal-2English
Titel16th International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2017
RedacteurenEdmund Durfee, Michael Winikoff, Kate Larson, Sanmay Das
Pagina's1705-1707
Aantal pagina's3
Volume3
ISBN van elektronische versie9781510855076
StatusPublished - 8 mei 2017
EvenementInternational Conference on Autonomous Agents and Multiagent Systems - Sao Paolo, Brazil
Duur: 8 mei 201712 mei 2017

Conference

ConferenceInternational Conference on Autonomous Agents and Multiagent Systems
Verkorte titelAAMAS 2017
LandBrazil
StadSao Paolo
Periode8/05/1712/05/17

Vingerafdruk

Duik in de onderzoeksthema's van 'Analysing Congestion Problems in Multi-agent Reinforcement Learning'. Samen vormen ze een unieke vingerafdruk.

Citeer dit