IoT-Based Cybersecurity Threat Detection Using Feature Se-lection and Dipper Throated Optimization Algorithm | ||||
| International Journal of Telecommunications | ||||
| Volume 05, Issue 02, July 2025, Page 1-22 PDF (1.27 MB) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/ijt.2025.387473.1106 | ||||
 View on SCiNiTO | ||||
| Authors | ||||
Ali Takieldeen   1; S Towfek2; Ahmed Zaki2; Nima Khodadadi3
| ||||
| 1Head of the Cyber Security Department, Faculty of Artificial Intelligence, Delta University for Science and Technology, Gamasa 35712, Mansoura, Egypt. | ||||
| 2Computer Science and Intelligent Systems Research Center, Blacksburg 24060, Virginia, USA | ||||
| 3Department of Civil and Environmental Engineering, University of California, Berkeley, CA, USA | ||||
| Abstract | ||||
| This work aims to investigate enhancing programs employed in threat identification associated with cybersecurity in the back-ground of the IoT in combination with the Dipper Throated Optimization (DTO) algorithm and Gradient Boosting. The increasing intricacy of information systems and a sharp increase in the usage of IoT devices would indicate that technology's . Facilitating the management of the arising challenges related to optimization problems in the field of cybersecurity is the playing ground of metaheuristic optimization algorithms based on the principles of natural sciences. These algorithms are well described in the literature, and this research carefully analyzes and deploys them to carry out feature selection with a focus on the IoT cybersecurity context. Specifically, they solve the typically tricky combinatorial optimization problem of binary optimization to feature selection to pick the most relevant features with the most negligible computation intelligently. Another increase in efficiency when applying the cybersecurity framework is when it is integrated with machine learning models. For the regression, the following approaches have been implemented: Gradient Boosting, CatBoost, and XGBoost. Besides, mean squared error (MSE) and the percentage of change in root mean squared error (RMSE) were used when comparing these models. The results of this research advance the scholarship of optimization in the context of IoT cybersecurity and hold practical implications for improved threat detection models' implementation in applications. Including DTO with Gradient Boosting enhances the attainment of high-quality cybersecurity threat detection in IoT, ensuring the value of speeding up modified interconnected systems. | ||||
| Keywords | ||||
| IoT cybersecurity attacks; Dipper Throated Optimization algorithm; Gradient Boosting; a | ||||
|
Statistics Article View: 86 PDF Download: 56 |
||||
