Prototype of a toolset for machine learning-enabled detection of vulnerabilities in CPS

The toolset will rely on machine learning to identify inputs triggering anomalous execution flows that characterize vulnerabilities. The analysis will be based on data collected at development time and in the field.