The Effectiveness of Machine Learning Techniques in the Detection of Multi-intrusion Attacks
Open AccessThe development of new adversarial technologies is making networks vulnerable to attacks. The increased sophistication of network intrusion techniques is also posing challenges for detection. Artificial intelligence (AI) based networks and cyber security management tools are becoming more available and gaining prominence. Machine learning algorithms are an essential part of AI systems. This praxis explores the effectiveness of machine learning techniques in the detection of intrusion attacks.In order to measure the effectiveness of ML techniques, the dataset went through a process. We removed from the original dataset features that exhibited no variability in measurement. To have equal weights for each class, sampling was performed. In order to account for the various distributions and factors, normalization was performed. Four techniques were found to perform well. The LightGBM model has a performance of 0.992 and a precision of 0.99. The Extra Gradient Boost (XGBoost) model and the extra three models both display an accuracy of 0.985 and a precision of 0.99. The CatBoost model has an accuracy of 0.975 and a precision of 0.98. The results we obtained surpass those documented in the literature for the same dataset and demonstrate comparability to results achieved with similar datasets. These satisfactory performances suggest a potential for these ML models to be deployed in cybersecurity management. Our study also shows that intrusion attacks of the “Overflow”, ‘PortScan” are more likely to be detected by the models than “TCP-SYN”. Areas for futures research concern the consideration of more intrusion types. Also, new revolutionary AI methods could be explored.
- 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.
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
DjahaFodja_gwu_0075A_16537.pdf | 2023-11-14 | Open Access |
|