Multi-Stage Cyber-Attack Prediction Using AI and Machine Learning Approaches

The PhD project will research future AI and machine learning methods to predict multi-stage cyber attacks. Modern cyber intrusion and attacks are often delivered and achieved in multi-stages. Among these multi-stage cyber-attacks, multi-stage malware attacks or multi-stage malware attacks bundled up with other types of attacks are very common. Famous examples include the 2017 NHS cyber attack, and the same families of malware attacked over 200,000 computers all over the world. The increase of these attacks becomes significant threats to the future cyberspace and digital world.

AI and machine learning methods, particularly deep learning, have proven effectiveness in automatically classifying objects and predict their groups and associations. This project will develop AI approaches to identify multi-stage malware attacks and how the different stages are associated within the same attack. Such knowledge will be used to forecast which malwares are likely to be seen in the near future given the malware detection data.

This PhD project will benefit from existing work (publications and patents) at Exeter and BT on developing novel machine learning and AI approaches to capture the multistage associations within a malware attack. Our existing models and data could be explored for an agile start. You will also learn to generate datasets and write high quality scientific publications, to promote further research and contribute to the cybersecurity research community.