Privacy for Data Sharing and Machine Learning

This project takes a fundamental approach to characterising and designing optimal privacy-preserving algorithms for a variety of applications in data sharing and machine learning. Important privacy frameworks such as differential privacy and information-theoretic Bayesian privacy are studied. Important data sharing applications such as official census data release and machine learning frameworks such as deep neural networks are targeted in this project.

PhD applicants with a solid mathematical and analytical background and motivation to learn new tools in engineering and computer science are welcome to apply. Prior experience in computer programming in Python is a plus. Prior research experience in masters degree and conference or journal publications is favourably considered.

Contact:

Parastoo Sadeghi p.sadeghi@unsw.edu.au