Streaming anomaly detection
Algorithms that identify unusual behavior and concept drift as data arrives, with attention to latency, robustness, and deployability.
PhD Student | School of Computing | Newcastle University
Ruitao Xue
I build resource-efficient and secure intelligence for the Internet of Things. My research connects streaming anomaly detection, TinyML, and lifelong learning so embedded devices can adapt at the edge without sacrificing reliability.
About Me
Edge intelligence that keeps learning after deployment.
I am a PhD student in Computing at Newcastle University. My work studies how tiny, networked devices can detect change, learn from streaming data, and remain dependable in messy real-world environments.
I am especially interested in practical algorithms that balance accuracy, energy cost, memory limits, and security needs for embedded sensing systems.
Research
Research focus
Tiny machine learning
Learning pipelines for constrained hardware such as Raspberry Pi, ESP32, and Arduino-class platforms.
Lifelong edge intelligence
Continual learning and autonomous agents for cyber-physical systems that need to adapt after deployment.
Publications
Selected publications
- ICDM OWAD 2025 STREAM-LAD: A Practical Streaming Lifelong Anomaly Detection Algorithm for IoT.
- EWSN EMERGE 2025 Energy-Efficient Change Point Detection Algorithm for Resource-Constrained Devices.
Education
Education
Newcastle University
PhD in Computing. Researching IoT intelligence, TinyML, anomaly detection, and agent-driven automation.
University of Glasgow
MSc in Computer Systems Engineering. Graduated with Distinction; GPA 17.8/22.
Beihang University (BUAA)
Bachelor of Engineering in Information Security. Graduated with Honors from ShenYuan Honors College.
Skills
Selected skills
Programming
Java, Python, PyTorch, NumPy, scikit-learn, C, C++, CUDA, Go, SQL
Embedded systems
Raspberry Pi, ESP32, Arduino, high-level and low-level development
Tooling
Linux, Bash, Zsh, LaTeX, Microsoft Office, Git
Languages
English, Chinese, Japanese
Collaborate