Machine-Learning Techniques to Protect Critical Infrastructure From Cybersecurity Incidents or Equipment Incidents

Securing our critical infrastructure is vital to national security. This presentation demonstrates techniques that can be used to detect cyberattacks or equipment failures. A typical industrial control system is used as an example, with data collected to monitor and control the system. Data science techniques to prepare the data for machine learning will be analyzed. Various machine-learning approaches used to detect anomalies due to cyberattacks or equipment failures will be demonstrated and their efficacy discussed.