Training agents for unknown logistics problems

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Autoren

  • Elisa Schmid
  • Matthias Becker
Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Titel des SammelwerksGECCO '23 Companion
UntertitelProceedings of the Companion Conference on Genetic and Evolutionary Computation
Seiten243-246
Seitenumfang4
ISBN (elektronisch)9798400701207
PublikationsstatusVeröffentlicht - 24 Juli 2023
Veranstaltung2023 Genetic and Evolutionary Computation Conference Companion: GECCO 2023 - Lisbon, Portugal
Dauer: 15 Juli 202319 Juli 2023

Abstract

A methodology on how to prepare agents to succeed on a priori unknown logistics problems is presented. The training of the agents is and can only be executed using a small number of test problems that are taken out of a broad class of generalized logistics problems. The developed agents are then evaluated on unknown instances of the problem class. This work has been developed in the context of last year’s AbstractSwarm Multi-Agent Logistics Competition. The most successful algorithms are presented, and additionally, all participating algorithms are discussed with respect to the features of the algorithms that contribute to their success. As a result, we conclude that such a broad variety of a priori unknown logistics problems can be solved efficiently if multiple different good working approaches are used, instead of trying to find one optimal algorithm. For the used test problems this method can undercut, trivial as well as non-trivial implementations, for example, algorithms based on machine learning.

ASJC Scopus Sachgebiete

Zitieren

Training agents for unknown logistics problems. / Schmid, Elisa; Becker, Matthias.
GECCO '23 Companion: Proceedings of the Companion Conference on Genetic and Evolutionary Computation. 2023. S. 243-246.

Publikation: Beitrag in Buch/Bericht/Sammelwerk/KonferenzbandAufsatz in KonferenzbandForschungPeer-Review

Schmid, E & Becker, M 2023, Training agents for unknown logistics problems. in GECCO '23 Companion: Proceedings of the Companion Conference on Genetic and Evolutionary Computation. S. 243-246, 2023 Genetic and Evolutionary Computation Conference Companion, Lisbon, Portugal, 15 Juli 2023. https://doi.org/10.1145/3583133.3590724
Schmid, E., & Becker, M. (2023). Training agents for unknown logistics problems. In GECCO '23 Companion: Proceedings of the Companion Conference on Genetic and Evolutionary Computation (S. 243-246) https://doi.org/10.1145/3583133.3590724
Schmid E, Becker M. Training agents for unknown logistics problems. in GECCO '23 Companion: Proceedings of the Companion Conference on Genetic and Evolutionary Computation. 2023. S. 243-246 doi: 10.1145/3583133.3590724
Schmid, Elisa ; Becker, Matthias. / Training agents for unknown logistics problems. GECCO '23 Companion: Proceedings of the Companion Conference on Genetic and Evolutionary Computation. 2023. S. 243-246
Download
@inproceedings{c982fb906404492f811165ca6657335b,
title = "Training agents for unknown logistics problems",
abstract = "A methodology on how to prepare agents to succeed on a priori unknown logistics problems is presented. The training of the agents is and can only be executed using a small number of test problems that are taken out of a broad class of generalized logistics problems. The developed agents are then evaluated on unknown instances of the problem class. This work has been developed in the context of last year{\textquoteright}s AbstractSwarm Multi-Agent Logistics Competition. The most successful algorithms are presented, and additionally, all participating algorithms are discussed with respect to the features of the algorithms that contribute to their success. As a result, we conclude that such a broad variety of a priori unknown logistics problems can be solved efficiently if multiple different good working approaches are used, instead of trying to find one optimal algorithm. For the used test problems this method can undercut, trivial as well as non-trivial implementations, for example, algorithms based on machine learning.",
keywords = "agent learning, competition, unknown logistics problems",
author = "Elisa Schmid and Matthias Becker",
year = "2023",
month = jul,
day = "24",
doi = "10.1145/3583133.3590724",
language = "English",
pages = "243--246",
booktitle = "GECCO '23 Companion",
note = "2023 Genetic and Evolutionary Computation Conference Companion : GECCO 2023 ; Conference date: 15-07-2023 Through 19-07-2023",

}

Download

TY - GEN

T1 - Training agents for unknown logistics problems

AU - Schmid, Elisa

AU - Becker, Matthias

PY - 2023/7/24

Y1 - 2023/7/24

N2 - A methodology on how to prepare agents to succeed on a priori unknown logistics problems is presented. The training of the agents is and can only be executed using a small number of test problems that are taken out of a broad class of generalized logistics problems. The developed agents are then evaluated on unknown instances of the problem class. This work has been developed in the context of last year’s AbstractSwarm Multi-Agent Logistics Competition. The most successful algorithms are presented, and additionally, all participating algorithms are discussed with respect to the features of the algorithms that contribute to their success. As a result, we conclude that such a broad variety of a priori unknown logistics problems can be solved efficiently if multiple different good working approaches are used, instead of trying to find one optimal algorithm. For the used test problems this method can undercut, trivial as well as non-trivial implementations, for example, algorithms based on machine learning.

AB - A methodology on how to prepare agents to succeed on a priori unknown logistics problems is presented. The training of the agents is and can only be executed using a small number of test problems that are taken out of a broad class of generalized logistics problems. The developed agents are then evaluated on unknown instances of the problem class. This work has been developed in the context of last year’s AbstractSwarm Multi-Agent Logistics Competition. The most successful algorithms are presented, and additionally, all participating algorithms are discussed with respect to the features of the algorithms that contribute to their success. As a result, we conclude that such a broad variety of a priori unknown logistics problems can be solved efficiently if multiple different good working approaches are used, instead of trying to find one optimal algorithm. For the used test problems this method can undercut, trivial as well as non-trivial implementations, for example, algorithms based on machine learning.

KW - agent learning

KW - competition

KW - unknown logistics problems

UR - http://www.scopus.com/inward/record.url?scp=85169034992&partnerID=8YFLogxK

U2 - 10.1145/3583133.3590724

DO - 10.1145/3583133.3590724

M3 - Conference contribution

AN - SCOPUS:85169034992

SP - 243

EP - 246

BT - GECCO '23 Companion

T2 - 2023 Genetic and Evolutionary Computation Conference Companion

Y2 - 15 July 2023 through 19 July 2023

ER -