Enhancing Intrusion Detection Using Dragonfly Algorithm-Based Feature Selection and Extra Trees for Classification

Intrusion Detection Systems (IDS) play a critical role in identifying cyber threats and securing network environments. This research proposes an optimized IDS by leveraging the Dragonfly Algorithm (DA) for feature selection and employing Extra Trees (ET) and K-Nearest Neighbors (KNN) for classification. The UNSW-NB15 dataset is used for evaluation, ensuring a robust and realistic assessment. The implementation, conducted in Python, demonstrates that ET, when combined with DA-based feature selection, achieves an accuracy of 100%, highlighting its effectiveness in distinguishing between normal and malicious traffic. The results confirm that intelligent feature selection significantly enhances IDS performance, reducing computational complexity while improving detection accuracy. This study contributes to the advancement of IDS methodologies, offering a highly efficient approach for real-world cybersecurity applications.