Using Federated Learning for Collaborative Intrusion Detection Systems

Neural networks have become cutting edge machine learning models for detecting network attacks. Traditional implementations provide fast and accurate predictions, but require centralised storage of labelled historical data for training. This solution is not always suitable for real-world application...

Full description

Saved in:
Bibliographic Details
Main Authors: Matteo Rizzato, Youssef Laarouchi, Christophe Geissler
Format: Article
Language:English
Published: International Institute of Informatics and Cybernetics 2023-06-01
Series:Journal of Systemics, Cybernetics and Informatics
Subjects:
Online Access:http://www.iiisci.org/Journal/PDV/sci/pdfs/SA291HE23.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850177566781997056
author Matteo Rizzato
Youssef Laarouchi
Christophe Geissler
author_facet Matteo Rizzato
Youssef Laarouchi
Christophe Geissler
author_sort Matteo Rizzato
collection DOAJ
description Neural networks have become cutting edge machine learning models for detecting network attacks. Traditional implementations provide fast and accurate predictions, but require centralised storage of labelled historical data for training. This solution is not always suitable for real-world applications, where regulatory constraints and privacy concerns hamper the collection of sensitive data into a single server. Federated Learning has recently been proposed as a framework for training a centralised model without the need to share data between different providers. We use the CICIDS2017 dataset provided by the Canadian Institute of Cybersecurity to demonstrate the benefits of Neural Networks-based Federated Learning for the detection of the most relevant types of network attacks. We conclude that a federated-trained neural network outperforms locally-trained models (at isoarchitecture) in terms of F1-score and False Negative detection ratio. Further, such model has a minor loss of performance and convergence rapidity compared to a model trained over a hypothetical centralised dataset.
format Article
id doaj-art-3103be35491c42dfa8ec5df8be5bd147
institution OA Journals
issn 1690-4524
language English
publishDate 2023-06-01
publisher International Institute of Informatics and Cybernetics
record_format Article
series Journal of Systemics, Cybernetics and Informatics
spelling doaj-art-3103be35491c42dfa8ec5df8be5bd1472025-08-20T02:18:57ZengInternational Institute of Informatics and CyberneticsJournal of Systemics, Cybernetics and Informatics1690-45242023-06-012132936Using Federated Learning for Collaborative Intrusion Detection SystemsMatteo RizzatoYoussef LaarouchiChristophe GeisslerNeural networks have become cutting edge machine learning models for detecting network attacks. Traditional implementations provide fast and accurate predictions, but require centralised storage of labelled historical data for training. This solution is not always suitable for real-world applications, where regulatory constraints and privacy concerns hamper the collection of sensitive data into a single server. Federated Learning has recently been proposed as a framework for training a centralised model without the need to share data between different providers. We use the CICIDS2017 dataset provided by the Canadian Institute of Cybersecurity to demonstrate the benefits of Neural Networks-based Federated Learning for the detection of the most relevant types of network attacks. We conclude that a federated-trained neural network outperforms locally-trained models (at isoarchitecture) in terms of F1-score and False Negative detection ratio. Further, such model has a minor loss of performance and convergence rapidity compared to a model trained over a hypothetical centralised dataset.http://www.iiisci.org/Journal/PDV/sci/pdfs/SA291HE23.pdf deep learningmulti-class classificationcybersecurityfederated learningneural networks
spellingShingle Matteo Rizzato
Youssef Laarouchi
Christophe Geissler
Using Federated Learning for Collaborative Intrusion Detection Systems
Journal of Systemics, Cybernetics and Informatics
deep learning
multi-class classification
cybersecurity
federated learning
neural networks
title Using Federated Learning for Collaborative Intrusion Detection Systems
title_full Using Federated Learning for Collaborative Intrusion Detection Systems
title_fullStr Using Federated Learning for Collaborative Intrusion Detection Systems
title_full_unstemmed Using Federated Learning for Collaborative Intrusion Detection Systems
title_short Using Federated Learning for Collaborative Intrusion Detection Systems
title_sort using federated learning for collaborative intrusion detection systems
topic deep learning
multi-class classification
cybersecurity
federated learning
neural networks
url http://www.iiisci.org/Journal/PDV/sci/pdfs/SA291HE23.pdf
work_keys_str_mv AT matteorizzato usingfederatedlearningforcollaborativeintrusiondetectionsystems
AT yousseflaarouchi usingfederatedlearningforcollaborativeintrusiondetectionsystems
AT christophegeissler usingfederatedlearningforcollaborativeintrusiondetectionsystems