Welcome to the InfoSci Platform
Reference Hub2
Comparative Analysis of Machine Learning-Based Algorithms for Detection of Anomalies in IIoT

Comparative Analysis of Machine Learning-Based Algorithms for Detection of Anomalies in IIoT

Bhupal Naik D. S., Venkatesulu Dondeti, Sivadi Balakrishna
Copyright: © 2022 |Volume: 12 |Issue: 1 |Pages: 55
ISSN: 2155-6377|EISSN: 2155-6385|EISBN13: 9781683182085|DOI: 10.4018/IJIRR.298647
Cite Article Cite Article

MLA

Bhupal Naik D. S., et al. "Comparative Analysis of Machine Learning-Based Algorithms for Detection of Anomalies in IIoT." IJIRR vol.12, no.1 2022: pp.1-55. http://doi.org/10.4018/IJIRR.298647

APA

Bhupal Naik D. S., Dondeti, V., & Balakrishna, S. (2022). Comparative Analysis of Machine Learning-Based Algorithms for Detection of Anomalies in IIoT. International Journal of Information Retrieval Research (IJIRR), 12(1), 1-55. http://doi.org/10.4018/IJIRR.298647

Chicago

Bhupal Naik D. S., Venkatesulu Dondeti, and Sivadi Balakrishna. "Comparative Analysis of Machine Learning-Based Algorithms for Detection of Anomalies in IIoT," International Journal of Information Retrieval Research (IJIRR) 12, no.1: 1-55. http://doi.org/10.4018/IJIRR.298647

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

With the enormous increase in data, anomaly detection plays a prominent role in the finer analysis process. IIoT represents the Industrial Internet of Things that at first chiefly alluded to a mechanical system whereby an enormous number of devices or machines are associated and synchronized using programming devices and third stage advancements in a machine-to-machine and Internet of Things, later an Industry 4.0. The data produced by multiple huge numbers of sensors are incredibly complicated, diverse, and massive in IIoT and is raw. These may contain anomalies which are needed to be identified for better data analysis. In this research, we compare the Machine Learning algorithms of classification for detecting anomalies. The algorithms being compared here are Random Forest (RF), Logistic Regression (LR), Light Gradient Boosting Machine (LightGBM), Decision Trees (DT), K Nearest Neighbors (KNN). Three IIOT benchmark datasets were taken into consideration for analysis. The results have shown that RF has outperformed other algorithms used for the detection of anomalies in IIoT data.