"Intrusion Alert: System Uses Machine Learning, Curiosity-Driven ‘Honeypots’ to Stop Cyber Attackers"
The FBI has brought further attention to the increased targeting of government systems and networks in cyberattacks. In an effort to help stop these attacks, researchers at Purdue University developed a detection system called LIDAR (lifelong, intelligent, diverse, agile, and robust). The system operates through the use of supervised machine learning, unsupervised machine learning, and rule-based learning. The implementation of these types of machine learning allows LIDAR to detect anomalies in the system, compare detected abnormalities to known attack templates, and determine the validity of a potential attack. LIDAR also uses a honeypot to attract attackers without allowing them to enter the system. This article continues to discuss the components and capabilities of Purdue's LIDAR system.