Overview
Hadi M. Dolatabadi is a Research Fellow in machine learning at the University of Melbourne node of ARC Centre of Excellence for Automated Decision-Making & Society (ADM+S).
Hadi has nearly finished his Ph.D. at the School of Computing and Information Systems at the University of Melbourne, focusing on robustness in deep learning. His thesis examines current notions of robustness in neural networks and challenges them from novel perspectives.
Hadi has expertise in different areas of machine learning, ranging from adversarial robustness and coreset selection to generative modeling (normalizing flows, GANs, and diffusion models). In a nutshell, Hadi enjoys formulating various problems through the lens of generative modeling and bringing a statistical perspective to them. He has a solid track record with publications in world-class AI conferences such as NeurIPS, ECCV, and AISTATS. During his Ph.D., Hadi also completed six months of a research internship at Amazon Science.
Hadi will contribute to ADM+S Centre’s Machines Research Program. His research entails developing algorithms that enable a systematic treatment of bias and unfairness in AI and automated decision-making.