Enhancing infrastructure security with robust machine learning

Many tools and techniques for infrastructure security are continuously enhanced and improved. Improvements are – for example – based on increasing intelligence and analysis of relevant data. There are encouraging results and significant efforts in exploring advanced AI and Machine Learning (ML) techniques to enhance the quality and capabilities of many security technologies. This work is typically based on the core principle of training solutions by using high-quality data sets; yet many challenges remain. For example, it remains challenging to collect appropriate data sets that are sufficiently representative for the problems at hand. In addition, in order to keep up with an ever-changing environment, it is more than worthwhile to investigate the power and possibilities of security solutions powered by new and evolving algorithms in the ML space. Lastly, prototypical solutions cannot be deployed in practice unless they are equipped to face strong and adaptive attackers and adversaries. These are some of the important inroads that will be pursued over the course of this PhD project.