Enhanced Detection of Network Intrusions Using the C4.5 Decision Tree Algorithm in Complex Cybersecurity Environments

Abstract

The detection of network intrusions is increasingly important as cyberattacks become more sophisticated and frequent. This study explores the efficacy of the C4.5 decision tree algorithm in detecting network intrusions within complex cybersecurity environments. Utilizing the NSL-KDD dataset, an improved version of the KDD Cup 1999 dataset, we trained and tested our model to ensure comprehensive and reliable results. Our methodology included meticulous data preprocessing steps, such as cleaning, normalization, and categorical encoding, followed by model building and performance evaluation. Results indicate that the C4.5 algorithm effectively classifies network activities with high accuracy (89.25%), precision (86.50%), recall (90.75%), and F1-score (88.57%). The confusion matrix analysis further validates the model’s robustness, highlighting high true positive and true negative rates. This research significantly contributes to the development of robust intrusion detection systems, offering a scalable solution for real-world network security challenges. By addressing the evolving nature of cyber threats, this study provides actionable insights for network security practitioners and sets a foundation for future research in enhancing intrusion detection capabilities.