A masking-based federated singular value decomposition method for anomaly detection in industrial internet of things
by Olena Hordiichuk-Bublivska; Halyna Beshley; Natalia Kryvinska; Mykola Beshley
International Journal of Web and Grid Services (IJWGS), Vol. 19, No. 3, 2023

Abstract: The industrial internet of things (IIoT) is a flexible and scalable manufacturing system that can collect and analyse data from sensors based on machine learning, cloud, and edge computing. Recommendation systems can identify patterns in big data and reduce irrelevant data, with the singular value decomposition (SVD) algorithm being commonly used. Based on the found regularities in the data, it is possible to predict the most probable future events, such as emergency shutdowns of equipment, the occurrence of emergencies, etc. This paper explores the SVD method for anomaly detection in IIoT and proposes the federated singular value decomposition (FedSVD) method, which better protects large-scale IIoT data privacy. Results show FedSVD has greater accuracy and duration of calculations. A masking-based FedSVD method is proposed for anomaly detection and data protection. Choosing the optimal algorithm for IIoT and recommendation systems can automate the processing of critical parameters and improve efficiency.

Online publication date: Mon, 18-Sep-2023

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Web and Grid Services (IJWGS):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com