Please use this identifier to cite or link to this item:
Publication type: Conference paper
Type of review: Peer review (publication)
Title: Improving sample efficiency and multi-agent communication in RL-based train rescheduling
Authors: Roost, Dano
Meier, Ralph
Huschauer, Stephan
Nygren, Erik
Egli, Adrian
Weiler, Andreas
Stadelmann, Thilo
et. al: No
DOI: 10.21256/zhaw-19978
Proceedings: Proceedings of the 7th SDS
Conference details: 7th Swiss Conference on Data Science, Lucerne, Switzerland, 26 June 2020
Issue Date: 26-Jun-2020
Publisher / Ed. Institution: IEEE
Language: English
Subjects: Multi-agent deep reinforcement learning
Subject (DDC): 006: Special computer methods
Abstract: We present preliminary results from our sixth placed entry to the Flatland international competition for train rescheduling, including two improvements for optimized reinforcement learning (RL) training efficiency, and two hypotheses with respect to the prospect of deep RL for complex real-world control tasks: first, that current state of the art policy gradient methods seem inappropriate in the domain of high-consequence environments; second, that learning explicit communication actions (an emerging machine-to-machine language, so to speak) might offer a remedy. These hypotheses need to be confirmed by future work. If confirmed, they hold promises with respect to optimizing highly efficient logistics ecosystems like the Swiss Federal Railways railway network.
Fulltext version: Accepted version
License (according to publishing contract): Licence according to publishing contract
Departement: School of Engineering
Organisational Unit: Institute of Applied Information Technology (InIT)
Appears in collections:Publikationen School of Engineering

Files in This Item:
File Description SizeFormat 
2020_Roost-etal_RL-based-train-rescheduling_SDS2020.pdfAccepted Version347.87 kBAdobe PDFThumbnail

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.