<![CDATA[The development of the Internet of Things has encouraged the use of connected devices across various sectors, including healthcare, energy, transportation, and household environments. However, the heterogeneous and dynamic nature of IoT networks, along with their reliance on wireless communication, makes them vulnerable to various cyber threats. This study aims to develop and evaluate an intrusion detection model for multiclass classification in IoT networks using the ToN_IoT dataset. The models used in this study are Random Forest and XGBoost, tested under several scenarios, including baseline models, models with hyperparameter tuning, and the application of SMOTENC to XGBoost to address class imbalance. The research stages include exploratory data analysis, data cleaning, feature selection, data splitting, feature encoding, data balancing, modeling, and evaluation using accuracy, precision, recall, F1-score, training time, and confusion matrix. The results show that all models achieved high performance, with accuracy and F1-score values above 99%. The best performance was obtained by XGBoost with SMOTENC, achieving an accuracy of 99.59% and an F1-score of 99.59%.]]>
Copyrights © 2026
