Comprehensive Analysis and Detection of Multi-Step Cyber Attacks: Insights from the Multi-Step Cyber-Attack Dataset
Atul Kumar, Kalpna Guleria · 2024
The rapid growth in the complexity of cyber threats has made using advanced detection techniques necessary, especially for multi-step cyber-attacks involving wellorchestrated operations aimed at infiltrating systems. This research study proposes a detailed analysis and detection approach for multi-step cyber-attacks with the Multi-Step Cyber-Attack Dataset. Rapidly use three machine learning algorithms, Random Forest, SVM, and KNN, to detect and examine attack patterns. Other evaluation metrics include accuracy, precision, recall, and the F1-score. All these measures are used to ensure comprehensiveness regarding algorithm performance assessment. The Random Forest method proved most effective and showed excellent accuracy and robustness in identifying complex, multi-step attack sequences. The research also highlights the effectiveness of Random Forest in cyberattack detection, underscoring rich datasets much like MultiStep Cyber-Attack Dataset, further improving cybersecurity research and practice.