Revisiting Gradient Boosting-Based Approaches for Learning Imbalanced Data: A Case of Anomaly Detection on Power Grids

Louk, Maya Hilda Lestari and Tama, Bayu Adhi (2022) Revisiting Gradient Boosting-Based Approaches for Learning Imbalanced Data: A Case of Anomaly Detection on Power Grids. Big Data Cognitive Computing, 6 (2). 41/1-9. ISSN 2504-2289

[thumbnail of BDCC-06-00041.pdf] PDF
BDCC-06-00041.pdf - Accepted Version
Restricted to Repository staff only

Download (484kB) | Request a copy
Official URL / DOI: https://www.mdpi.com/2504-2289/6/2/41

Abstract

Gradient boosting ensembles have been used in the cyber-security area for many years; nonetheless, their efficacy and accuracy for intrusion detection systems (IDSs) remain questionable, particularly when dealing with problems involving imbalanced data. This article fills the void in the existing body of knowledge by evaluating the performance of gradient boosting-based ensembles, including gradient boosting machine (GBM), extreme gradient boosting (XGBoost), LightGBM, and CatBoost. This paper assesses the performance of various imbalanced data sets using the Matthew correlation coefficient (MCC), area under the receiver operating characteristic curve (AUC), and F1 metrics. The article discusses an example of anomaly detection in an industrial control network and, more specifically, threat detection in a cyber-physical smart power grid. The tests’ results indicate that CatBoost surpassed its competitors, regardless of the imbalance ratio of the data sets. Moreover, LightGBM showed a much lower performance value and had more variability across the data sets.

Item Type: Article
Uncontrolled Keywords: imbalance learning; oversampling; anomaly detection; gradient boosting ensembles; power grid; MWMOTE
Subjects: T Technology > T Technology (General)
Divisions: Faculty of Engineering > Department of Informatic
Depositing User: MAYA HILDA LESTARI LOUK
Date Deposited: 18 Apr 2022 01:49
Last Modified: 18 Apr 2022 01:49
URI: http://repository.ubaya.ac.id/id/eprint/41773

Actions (login required)

View Item View Item