A CatBoost-Based Approach for High-Accuracy Botnet Detection

Abstract

The rising prevalence of network botnet attacks poses a significant threat to online security. Compromised networks controlled by malicious entities can perpetrate harm, including distributed denial of service attacks and data theft. In this study, we introduce a method to detect these botnets using the CatBoostClassifier. By analyzing network traffic for suspicious patterns, our system efficiently identifies potential botnet activities. Utilizing the CTU-13 dataset, we achieved an impressive 99.8699% accuracy, underscoring the efficacy of our approach. This research offers valuable insights into botnet attack detection and presents a robust solution for enhancing network security.