Robert Flood

Background

About Robert Flood

Hi there, I'm Rob Flood. I'm a PhD student at the University of Edinburgh investigating synthetic data generation applied to network traffic, particularly for intrusion detection. I have published in top conferences/workshops, including a unanimous win of a best paper award at EuroS & P for my work on experiment bias in NIDS datasets. I'm passionate about netsec, machine learning, secure programming and vulnerability/exploit discovery. I'm in the process of finishing my PhD thesis about enhancing the trustworthiness of ML-based methods applied to network data. My thesis combines the above work on security dataset analysis as well as research into security applications of neuro-symbolic methods, particularly formal verification of ML models. Whilst I wrap that up, I'm also working on improving ML-based vulnerability detectors for code, investigating both transformer-based and graph-based approaches to vulnerability discovery. I've graduated from the University of Edinburgh with an MSc in Computer Science with a distinction. My primary topics of study were Computer Security and Machine Learning/Artificial Intelligence. Furthermore, my dissertation - detailed below - focused on the intersection between these two topics. My Bachelor's degree is a BA from Trinity College Dublin in Mathematics with emphasis on Computer Science and Theoretical Physics. I've also worked as an Assistant Researcher with the University of Edinburgh, working on formal methods applied to Android contact tracing apps. Github: https://github.com/glo-fi/