The HK Laureate Forum offers a unique and prestigious platform to engage with world-leading minds, and I am excited about the opportunity to participate as a postdoctoral researcher in AI Trust and Security at the University of Manchester. My research focuses on fraud detection and understanding crime patterns on the blockchain using tools from financial mathematics, statistics, and network analysis. As digital crimes grow in complexity, interdisciplinary collaboration is essential to tackle global challenges in data security and trust. The Forum will allow me to share my current findings, including how advanced analytical methods can uncover fraud and anomalies in blockchain systems, while gaining insights from other disciplines. I am particularly eager to exchange ideas with fellow early-career researchers and Nobel Laureates, learning from their approaches to complex problems and building long-term collaborative networks. This experience will not only deepen my understanding of mathematical applications in emerging digital ecosystems but also strengthen my commitment to research with real-world impact. The Forum's emphasis on innovation, global dialogue, and cross-disciplinary exchange aligns perfectly with my academic vision and will be instrumental in shaping the next phase of my research career.

University of Manchester

University of Manchester

Nick_YZ_reference1.pdf

My current research lies at the intersection of AI Trust and Security, with a strong emphasis on fraud detection and crime pattern analysis within blockchain ecosystems. Building on my background in anomaly detection, I develop and apply advanced AI models and mathematical techniques to uncover illicit behaviours in digital financial systems. A core component of my work involves modelling blockchain networks as complex graphs, where nodes represent cryptocurrency wallets and edges represent transactional relationships. Leveraging graph neural networks, extreme value theory, and unsupervised learning methods, I identify outliers and suspicious structures indicative of fraudulent activity.
I also implement large language models and text mining techniques to analyse social platforms and web data, enriching on-chain analysis with off-chain behavioural signals. Tools such as R Shiny enable me to develop interactive dashboards for visualising and interpreting blockchain forensic data in real time.
My research integrates explainable AI with robust statistical methods to improve transparency, accountability, and trust in digital finance systems. This interdisciplinary work aims to develop practical tools that can support regulators, law enforcement, and financial institutions in monitoring, investigating, and preventing new and evolving forms of cyber-enabled crime within blockchain networks.