Siamak Layeghy

I supervise postgraduate research (MPhil/PhD) at the intersection of AI/ML, security and networking. Projects are typically hands-on and data-driven, with an emphasis on building deployable methods and producing publishable outcomes. Current directions include: (i) applied machine learning for security and networked systems; (ii) large language and foundation models for security analytics and automation; and (iii) adversarially robust and continual learning for resilient detection in dynamic environments.

Students in my group typically contribute to open research artefacts alongside publications. Applicants who do well usually have strong programming skills (e.g., Python/C++), a solid foundation in either ML or networking, and a willingness to engage deeply with sound experimental methodology and reproducible research.

About Me

I am a researcher at The University of Queensland working on practical AI/ML for security and networked systems. My research focuses on robust, scalable techniques for detecting and understanding malicious behaviour in modern networks and computing environments, and on methods that remain reliable under evolving threats and shifting data distributions. Current themes in my work include large language and foundation models for security analytics and automation, adversarially robust and continual learning, and explainable machine learning to support trustworthy operational use.