Identifying and Correlating Anomalies in Internetwide Scan
Identifying and correlating anomalies in internet-wide network scan data using machine learning.
Summary
Network Scanning Scale
Internet-wide scanning generates massive datasets. Identifying meaningful anomalies in this data is challenging without machine learning approaches.
Anomaly Detection
The presentation covers techniques for identifying anomalies in scan data and correlating them across scans. Understanding patterns in global internet scans reveals infrastructure vulnerabilities.
Machine Learning Application
ML models trained on network scan data can identify unusual patterns, misconfigurations, and potentially vulnerable systems at scale.
Applications
Identifying anomalies in internet-wide scans enables discovery of vulnerable services and infrastructure before attackers do.
Related talks
Fortinet·6 min
Engineers explore architecture and security mechanisms for FortiSIEM Machine Learning: Bipartite Graph Edge Anomaly Detection | Security Operations.
Strange Loop Conference·41 min
Presenters detail vulnerability discovery and exploitation in "Scalable Anomaly Detection (with Zero Machine Learning)" by Arthur Gonigberg.
Car Hacking Village·19 min
Ravi Rajput
Introduces a sequence-learning approach that analyzes inter-signal correlations in raw CAN bus payloads to detect stealthy contextual automotive attacks.