<![CDATA[This research aims to enhance the performance of the Random Forest algorithm in classifying big data within the Internet of Things (IoT) domain, specifically for detecting botnet attacks. The study utilizes the N-BaIoT dataset, comprising 150,000 instances of IoT network traffic categorized into normal and anomalous (botnet) data. To optimize classification outcomes, a preprocessing technique—the “remove frequent values” filter—is applied to reduce redundancy and improve computational efficiency. Model performance is evaluated using accuracy, precision, recall, and F1-score. Experimental results demonstrate that this filter improves classification accuracy from 99.976% to 99.998%, with precision, recall, and F1-score all reaching 1.000. Cross-validation was conducted to ensure the robustness of these results. These findings suggest that even lightweight preprocessing techniques can significantly enhance machine learning performance in IoT big data classification tasks. ]]>
Copyrights © 2025
