Federated Learning for Anomaly Detection in Maritime Movement Data

Anita Graser; Axel Weißenfeld; Clemens Heistracher; Melitta Dragaschnig; Peter Widhalm

doi:10.1109/MDM61037.2024.00030

Federated Learning for Anomaly Detection in Maritime Movement Data

Anita Graser (Autor:in und Vortragende:r), Axel Weißenfeld, Clemens Heistracher, Melitta Dragaschnig, Peter Widhalm

craftworks GmbH

Publikation: Beitrag in Buch oder Tagungsband › Vortrag mit Beitrag in Tagungsband › Begutachtung

Abstract

This paper introduces M3fed, a novel solution for federated learning of movement anomaly detection models. This innovation has the potential to improve data privacy and reduce communication costs in machine learning for movement anomaly detection. We present the novel federated learning (FL) strategies employed to train M3fed, perform an example experiment with maritime AIS data, and evaluate the results with respect to communication costs and FL model quality by comparing classic centralized M3 and the new federated M3fed.

Originalsprache	Englisch
Titel	Proceedings of MDM 2024
Seiten	77-82
DOIs	https://doi.org/10.1109/MDM61037.2024.00030
Publikationsstatus	Veröffentlicht - 2024
Veranstaltung	25th IEEE International Conference on Mobile Data Management - Brussels, Belgien Dauer: 24 Juni 2024 → 27 Juni 2024 https://mdm2024.github.io/

Konferenz

Konferenz	25th IEEE International Conference on Mobile Data Management
Kurztitel	MDM 2024
Land/Gebiet	Belgien
Stadt	Brussels
Zeitraum	24/06/24 → 27/06/24
Internetadresse	https://mdm2024.github.io/

Research Field

Multimodal Analytics
Transport Optimisation and Energy Logistics

Zugriff auf Dokument

10.1109/MDM61037.2024.00030

Diese Publikation zitieren

@inproceedings{765da39dc4784daf94b59a6c7b4d99d8,

title = "Federated Learning for Anomaly Detection in Maritime Movement Data",

abstract = "This paper introduces M3fed, a novel solution for federated learning of movement anomaly detection models. This innovation has the potential to improve data privacy and reduce communication costs in machine learning for movement anomaly detection. We present the novel federated learning (FL) strategies employed to train M3fed, perform an example experiment with maritime AIS data, and evaluate the results with respect to communication costs and FL model quality by comparing classic centralized M3 and the new federated M3fed.",

author = "Anita Graser and Axel Wei{\ss}enfeld and Clemens Heistracher and Melitta Dragaschnig and Peter Widhalm",

year = "2024",

doi = "10.1109/MDM61037.2024.00030",

language = "English",

isbn = "979-8-3503-7455-1",

pages = "77--82",

booktitle = "Proceedings of MDM 2024",

note = "25th IEEE International Conference on Mobile Data Management, MDM 2024 ; Conference date: 24-06-2024 Through 27-06-2024",

url = "https://mdm2024.github.io/",

}

TY - GEN

T1 - Federated Learning for Anomaly Detection in Maritime Movement Data

AU - Weißenfeld, Axel

AU - Heistracher, Clemens

AU - Dragaschnig, Melitta

AU - Widhalm, Peter

A2 - Graser, Anita

PY - 2024

Y1 - 2024

N2 - This paper introduces M3fed, a novel solution for federated learning of movement anomaly detection models. This innovation has the potential to improve data privacy and reduce communication costs in machine learning for movement anomaly detection. We present the novel federated learning (FL) strategies employed to train M3fed, perform an example experiment with maritime AIS data, and evaluate the results with respect to communication costs and FL model quality by comparing classic centralized M3 and the new federated M3fed.

AB - This paper introduces M3fed, a novel solution for federated learning of movement anomaly detection models. This innovation has the potential to improve data privacy and reduce communication costs in machine learning for movement anomaly detection. We present the novel federated learning (FL) strategies employed to train M3fed, perform an example experiment with maritime AIS data, and evaluate the results with respect to communication costs and FL model quality by comparing classic centralized M3 and the new federated M3fed.

U2 - 10.1109/MDM61037.2024.00030

DO - 10.1109/MDM61037.2024.00030

M3 - Conference Proceedings with Oral Presentation

SN - 979-8-3503-7455-1

SP - 77

EP - 82

BT - Proceedings of MDM 2024

T2 - 25th IEEE International Conference on Mobile Data Management

Y2 - 24 June 2024 through 27 June 2024

ER -

Federated Learning for Anomaly Detection in Maritime Movement Data

Abstract

Konferenz

Research Field

Zugriff auf Dokument

Fingerprint

Diese Publikation zitieren