Hao Tan (谭浩)

I am a Ph.D. student at the University of Leeds, under the supervision of Dr. Nabi Omidvar and co-supervision at SUSTech. The vision of my research may contain three parts: how to analyze the neural network, how to design the neural network, and how to deploy the neural network.
Firstly, I will use Information Theory and Feature Visualization to interpret the behaviour of the neural network or find the boundary of it. Secondly, the knowledge we discover can be used as a prior to design the neural network. Thirdly, we will extend the designed neural network into a search space and use Neural Architecture Search to further improve its performance, or for deployment purposes when considering multiple objectives. What is more, my long-term goal is to design a general neural network that can process multi-modal information, such as texts and images, at the same time.

After my graduation, I continued to do the academic research for neural architecture search as a Research Assitant at SUSTech. During this one year, I submitted my papers into journals and handled the responses. At the same time, I was the key member and responsible for developing neural architecture search algorithms under two research giants.