Rate My Professor Sky Miao

Dr Sky Miao is a Lecturer in Computing and Information Technology in the School of Computer and Information Sciences at the University of Newcastle, Australia, within the College of Engineering, Science and Environment. She received her PhD from Swinburne University of Technology in 2021. Her research interests focus on security and privacy in machine learning, automatic speech recognition, big data analysis, and network security. She is a member of the Centre for Applied and Responsible AI. Dr Miao commenced her lecturing position at the University of Newcastle in 2021. Previously, she was a Postdoctoral Research Associate at Swinburne University of Technology from August 2021 to February 2022, working on the CRT Trustworthy Machine Learning project with Data61, Monash University, and the University of Melbourne. From July 2019 to December 2021, she served as a Sessional Lecturer, Tutor, and Instructor at Swinburne University, teaching IT Security (COS30015) to approximately 250 students, and as instructor and course designer for IoT Security and Cloud Computing (CC5904) at James Cook University. She is the Program Convenor for the Bachelor of Computing (Honours) and has mentored year-10 students in a cybersecurity research project through the P-Tech program since 2021.

Dr Miao has an extensive publication record in top-tier venues. Key journal articles include 'Machine Learning-based Cyber Attacks Targeting on Controlled Information: A Survey' in ACM Computing Surveys (2021), 'The audio auditor: user-level membership inference in Internet of Things voice services' in Proceedings on Privacy Enhancing Technologies (2021), 'A Comprehensive Review on Graph-Based Anomaly Detection: Approaches for Intrusion Detection' in Applied Sciences (2026), and 'A survey of coverage-guided greybox fuzzing with deep neural models' in Information and Software Technology (2025). Prominent conference papers are 'No-Label User-Level Membership Inference for ASR Model Auditing' at ESORICS 2022, 'BadFU: Backdoor Federated Learning through Adversarial Machine Unlearning' (2025), 'Prompting Instability: An Empirical Study of LLM Robustness in Code Vulnerability Detection' (2026), and multiple 2024 works on optimizing large language models such as 'OptLLM: Optimal Assignment of Queries to Large Language Models' and 'CPLS: Optimizing the Assignment of LLM Queries'. In 2023, she led the CSIRO Next Generation Graduates Program grant 'Developing Digital Capabilities to Support the Aged Care Sector'. As principal supervisor, she oversaw a PhD completion in 2025 on 'Optimizing large language model utilization through scheduling strategies'. Her research is classified 80% in cybersecurity and privacy not elsewhere classified and 20% in machine learning not elsewhere classified.