To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Parameter Optimization in a Convolutional Neural Network for Cyberattacks Detection |
|
Author |
Maymouna M. Shbail, Khaled Batiha, and Wafa Alshatafat |
| Citation |
Vol. 26 No. 5 pp. 1-10 |
|
Abstract |
The rapid evolution of cyber threats poses significant challenges to traditional detection methods, which often lag in identifying advanced attacks and suffer from high false-positive rates. This paper integrated advanced optimization techniques and machine learning to enhance the accuracy and efficiency of intrusion detection systems (IDS). Specifically, it leverages the Jaya optimization algorithm to tune the hyperparameters of convolutional neural networks (CNNs) and support vector machines (SVMs) to detect cyber-attacks, particularly in innovative grid environments. Using the dataset UNSW-NB 15, the accuracy ratio was 99.7%. This paper addressed issues such as feature filtration and model interpretability. The proposed framework, JAYA-CNN-SVM, significantly improved classification accuracy and achieved robust detection across various attack types. The results confirm the effectiveness of the Jaya optimization in overcoming the limitations of traditional tuning methods, representing a step forward in reliable and real-time cybersecurity defenses. |
|
Keywords |
Convolutional Neural Networks, Feature Filtration, JAYA algorithm, Support Vector Machines. |
|
URL |
