research
Adversarial machine learning for tabular data, developing work in privacy-preserving healthcare AI, and collaborations in energy systems.
My research centres on adversarial machine learning for tabular data, with a focus on imperceptibility, attack evaluation and model robustness. My postdoctoral work explores privacy-preserving, distributed healthcare AI. I also contribute to collaborations on electricity price forecasting and battery storage optimisation.
Adversarial machine learning for tabular data
This is the main foundation of my research. My PhD work connects three questions: what makes a tabular attack imperceptible, how existing attacks compare, and how to generate attacks that better preserve the data distribution. Together, these studies examine how to evaluate model vulnerabilities using realistic changes to structured data.
Explore my adversarial ML research.
Healthcare AI: developing research direction
Through my postdoctoral work on AI-To-Data at QUT, I am developing expertise in adaptable, scalable learning across institutions and medical-data modalities. My responsibilities include training and evaluation pipelines and validation across distributed nodes, with a focus on privacy-preserving collaboration.
Energy systems: active collaborations
Alongside my main research, I collaborate on deep learning for electricity price forecasting under volatility in the Australian National Electricity Market. These collaborations also cover multi-agent reinforcement learning and cross-scale optimisation of Battery Energy Storage Systems (BESS), coordinating household and community operation with wider grid objectives.