A Cost-Sensitive Machine Learning Model With Multitask Learning for Intrusion Detection in IoT - Paper summary

Aug 5, 2026

A cost-sensitive machine learning approach is presented for intrusion detection in the Internet of Things (IoT), focusing on the limitations of conventional ML methods for low-frequency intrusions and for attacks not belonging to known categories. The work combines cost-sensitive learning and multitask learning within a hybrid model: an autoencoder performs feature extraction and an SVM detects intrusions. Low-distributed intrusions are addressed by enhancing the hinge loss layer for class imbalance. Unknown attacks are handled by formulating the SVM as a multitask problem, validated on UNSW-NB15 and BoT-IoT datasets with strong improvements in recall, precision, and F1-score.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages12

Reading Time30 min

AccessFree to download

Abstract and problem statement

Proposed hybrid model and learning strategies

Autoencoder feature extraction

Cost-sensitive hinge loss for class imbalance

Multitask formulation for unknown attacks

Experimental setup and results

Datasets used

Performance metrics

University of Groningen A Cost-Sensitive Machine Learning Model With Multitask Learning for Intrusion Detection in IoT Telikani, Akbar; Rudbardeh, Nima Esmi; Soleymanpour, Shiva; Shahbahrami, Asadollah; Shen, Jun; Gaydadjiev, Georgi; Hassanpour, Reza Published in: Ieee transactions on industrial informatics DOI: 10.1109/TII.2023.3314208 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2024 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Telikani, A. , Rudbardeh, N. E. , Soleymanpour, S. , Shahbahrami, A. , Shen, J. , Gaydadjiev, G. , & Hassanpour, R. (2024) . A Cost-Sensitive Machine Learning Model With Multitask Learning for Intrusion Detection in IoT. Ieee transactions on industrial informatics, 20(3), 3880-3890. [https://doi.org/10.1109/TII.2023.3314208](https://doi.org/10.1109/TII.2023.3314208) Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons) . The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: [https://www.rug.nl/library/open-access/self-archiving-pure/taverne](https://www.rug.nl/library/open-access/self-archiving-pure/taverne)amendment. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/U MCG research database (Pure): [http://www.rug. nl/research/portal. For technical reasons the](http://www.rug. nl/research/portal. For technical reasons the)[ ](http://www.rug. nl/research/portal. For technical reasons the)[number of authors shown on this cover page is limited to 10 maximum.](number of authors shown on this cover page is limited to 10 maximum.) Download date: 01-01-2026 3880 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 20, NO. 3, MARCH 2024 A Cost-Sensitive Machine Learning Model With Multitask Learning for Intrusion Detection in IoT Akbar Telikani , Graduate Student Member, IEEE, Nima Esmi Rudbardeh , Graduate Student Member, IEEE, Shiva Soleymanpour , Asadollah Shahbahrami , Jun Shen , Senior Member, IEEE, Georgi Gaydadjiev , and Reza Hassanpour Abstract—A problem with machine learning (ML) techniques for detecting intrusions in the Internet of Things (IoT) is that they are ineffective in the detection of lowfrequency intrusions. In addition, as ML models are trained using speciﬁc attack categories, they cannot recognize unknown attacks. This article integrates strategies of costsensitive learning and multitask learning into a hybrid ML model to address these two challenges. The hybrid model consists of an autoencoder for feature extraction and a support vector machine (SVM) for detecting intrusions. In the cost-sensitive learning phase for the class imbalance problem, the hinge loss layer is enhanced to make a classiﬁer strong against low-distributed intrusions. Moreover, to detect unknown attacks, we formulate the SVM as a multitask problem. Experiments on the UNSW-NB15 and BoT-IoT datasets demonstrate the superiority of our model in terms of recall, precision, and F1-score averagely 92.2%, 96.2%, and 94.3%, respectively, over other approaches. Index Terms—Deep learning (DL), Internet of things (IoT), intrusion detection, multitask learning, support vector machine (SVM) . NOMENCLATURE AE Autoencoder. CNN Convolutional neural networks. CSHL Cost-sensitive hi

FAQs

What are the two main challenges targeted for IoT intrusion detection?

The method addresses ineffective detection of low-frequency intrusions and the inability of models trained on specific attack categories to recognize unknown attacks.

How is the proposed hybrid model structured?

It uses an autoencoder for feature extraction and a support vector machine (SVM) for intrusion detection.

How does the approach improve detection under class imbalance?

During cost-sensitive learning, the hinge loss layer is enhanced so the classifier becomes stronger against low-distributed intrusions.

How does the model detect unknown attacks?

The SVM is formulated as a multitask problem to enable detection beyond known attack categories.

Aug 5, 2026

Upload by Ava Thompson