Anomaly Detection for Automated Cyber-attacks: Hybrid SVM ML-Based Approach

With the increasing sophistication and automation of cyber threats, traditional security measures are proving insufficient to safeguard critical systems and sensitive data. This paper presents a comprehensive approach to anomaly detection tailored for automated cyber-attacks. The proposed methodology integrates advanced machine learning algorithms, behavioral analysis, and anomaly detection techniques to identify deviations from normal system behavior. To apply the comprehensive approach to anomaly detection we have used the MAWILab dataset which covers diverse cyber threats, i.e., passive active attacks including spyware, worms, virus, backdoor Trojan, and many kinds emerging threats including but not limited to these attacks. In this paper, we are covering all such attacks with a depth analysis and an machine learning (ML)-based approach for detection and prediction based on the MAWILab dataset.