Cyber-Attack Detection in Critical Infrastructure Using Machine Learning
Open Access DepositedThe world is rapidly depending on the numerous services provided by the critical infrastructure units. However, the integration of Information Technology (IT) and Internet of Things (IoT) devices with the critical Infrastructure units for various purposes, including the management of the whole infrastructure, made it vulnerable to many malware and IoT-based cyber-attacks such as Ransomware and Mirai Botnet attacks, respectively. Considering the alarming circumstances faced by critical infrastructures such as Water Treatment Plants and Smart Grid Stations, this praxis has proposed a solution that utilizes Machine Learning (ML) methodology to proactively detect cyber-attacks before they can reach the critical infrastructure units. This praxis aims to test and train three ML models, Decision Tree (DT), Random Forest (RF), and Adaptive Boosting (AdaBoost), using the CIC-Malmem-2022, CIC-IoT-2023, CIC-IoT-2023-230, and CIC-IoT-2023-460 datasets.The methodology begins with the acquisition and preparation of these datasets, which replicate practical situations of malware and IoT-based attacks. The datasets were then visualized to understand and find important aspects of the datasets. Hence, it was found that the datasets are subjected to clean-up, encoded categorical columns, standardization, and normalization. After that, the datasets were segmented as testing and training sets, then the models were fitted and tested using performance parameters of accuracy, precision, F1 score, and recall. The findings from the praxis showed that all the three models have performed exceptionally well, achieving an accuracy greater than 99.00% across all datasets examined.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.