Dr Changyu Dong is a Senior Lecturer of Security in the School of Computing, Newcastle University. Before coming to Newcastle, he was a lecturer at the University of Strathclyde. He obtained his PhD from Imperial College London, on trust management in large distributed systems. His research focuses on privacy and security issues in data-driven systems. He has extensive experience in privacy-enhancing technologies, cryptographic protocol design and analysis, as well as secure computation. He led an EPSRC project “Practical Data-intensive Secure Computation: a Data Structural Approach”.
Machine learning often involves sensitive data and is used for critical tasks, hence maintaining and respecting trust, privacy, and sensitivity is of crucial importance. One of Changyu’s research interests is privacy-preserving machine learning. This has included work on designing various cryptographic protocols to compute on encrypted data, and/or mechanisms to ensure differential privacy. Another area of interest is on the robustness of machine learning, which has included work on designing mechanisms to defend against malicious attacks, such as adversarial examples or poisoning/backdoor attacks, that try to mislead or force machine learning models to make wrong decisions.