Advanced search
2 files | 3.66 MB Add to list

Sparse random neural networks for online anomaly detection on sensor nodes

Sam Leroux (UGent) and Pieter Simoens (UGent)
Author
Organization
Project
Abstract
Whether it is used for predictive maintenance, intrusion detection or surveillance, on-device anomaly detection is a very valuable functionality in sensor and Internet-of-things (IoT) systems. In this paper, we introduce a novel anomaly detection technique based on sparse, random neural networks. The sparsity in the model allows for a very efficient implementation on embedded or resource constrained hardware. Our approach supports continuous online learning where the model is deployed to the sensor device without any prior training. As new data becomes available, the model is updated and becomes better at detecting anomalies. We experimentally validate our approach on several default benchmark data sets in the visual domain as well as on industrial quality inspection and predictive maintenance tasks. We show that our approach achieves a very favorable trade-off between computational cost and anomaly detection accuracy.
Keywords
WEIGHTS, Anomaly detection, Continuous learning, TinyML, Edge AI, Sparse neural networks, Random neural networks

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.94 MB
  • Leroux S et al Accepted Manuscript.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 1.73 MB

Citation

Please use this url to cite or link to this publication:

MLA
Leroux, Sam, and Pieter Simoens. “Sparse Random Neural Networks for Online Anomaly Detection on Sensor Nodes.” FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, vol. 144, 2023, pp. 327–43, doi:10.1016/j.future.2022.12.028.
APA
Leroux, S., & Simoens, P. (2023). Sparse random neural networks for online anomaly detection on sensor nodes. FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, 144, 327–343. https://doi.org/10.1016/j.future.2022.12.028
Chicago author-date
Leroux, Sam, and Pieter Simoens. 2023. “Sparse Random Neural Networks for Online Anomaly Detection on Sensor Nodes.” FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 144: 327–43. https://doi.org/10.1016/j.future.2022.12.028.
Chicago author-date (all authors)
Leroux, Sam, and Pieter Simoens. 2023. “Sparse Random Neural Networks for Online Anomaly Detection on Sensor Nodes.” FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 144: 327–343. doi:10.1016/j.future.2022.12.028.
Vancouver
1.
Leroux S, Simoens P. Sparse random neural networks for online anomaly detection on sensor nodes. FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE. 2023;144:327–43.
IEEE
[1]
S. Leroux and P. Simoens, “Sparse random neural networks for online anomaly detection on sensor nodes,” FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, vol. 144, pp. 327–343, 2023.
@article{01H4G3RXWGNXX779S0T16PJPK9,
  abstract     = {{Whether it is used for predictive maintenance, intrusion detection or surveillance, on-device anomaly detection is a very valuable functionality in sensor and Internet-of-things (IoT) systems. In this paper, we introduce a novel anomaly detection technique based on sparse, random neural networks. The sparsity in the model allows for a very efficient implementation on embedded or resource constrained hardware. Our approach supports continuous online learning where the model is deployed to the sensor device without any prior training. As new data becomes available, the model is updated and becomes better at detecting anomalies. We experimentally validate our approach on several default benchmark data sets in the visual domain as well as on industrial quality inspection and predictive maintenance tasks. We show that our approach achieves a very favorable trade-off between computational cost and anomaly detection accuracy.}},
  author       = {{Leroux, Sam and Simoens, Pieter}},
  issn         = {{0167-739X}},
  journal      = {{FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE}},
  keywords     = {{WEIGHTS,Anomaly detection,Continuous learning,TinyML,Edge AI,Sparse neural networks,Random neural networks}},
  language     = {{eng}},
  pages        = {{327--343}},
  title        = {{Sparse random neural networks for online anomaly detection on sensor nodes}},
  url          = {{http://doi.org/10.1016/j.future.2022.12.028}},
  volume       = {{144}},
  year         = {{2023}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: