Optimisation has been an engineering tool shared by a large number of researchers. Traditional optimisation tools heavily rely on the mathematical model of the underlying problem. On the other hand, machine learning based optimisation approaches completely ignore the mathematical structure of the problem formulation. In this talk, we argue that model-free learning approaches and model-based optimisation approaches can work better together. In particular, we propose an actor-critic network structure to solve mixed integer non-linear programming problems with very low computational complexity. An application to wireless powered mobile edge computing system will be discussed as an example.
Associate Professor Yingjun Angela Zhang is currently with Department of Information Engineering, The Chinese University of Hong Kong. Her research interests include mainly wireless communications systems and smart power systems, in particular optimization techniques for such systems. She have served as the Chair of the Executive Editor Committee of the IEEE Transactions on Wireless Communications, and as an Associate Editor of the IEEE Transactions on Wireless Communications, IEEE Transactions on Communications, Security and Communications Networks (Wiley), and a Feature Topic in the IEEE Communications Magazine. She has served on the organizing committee of major IEEE conferences including ICC, GLOBECOM, SmartgridComm, VTC, CCNC, ICCC, MASS, etc. She is the Founding Chair of IEEE ComSoc Technical Committee on Smart Grid Communications. She was the co-recipient of the 2014 IEEE ComSoc APB Outstanding Paper Award, the 2013 IEEE SmartgridComm Best Paper Award, and the 2011 IEEE Marconi Prize Paper Award on Wireless Communications. She was the recipient of the Young Researcher Award from the Chinese University of Hong Kong in 2011. As the only winner from engineering science, she has won the Hong Kong Young Scientist Award 2006, conferred by the Hong Kong Institution of Science. Dr. Zhang is a Fellow of IET and a Distinguished Lecturer of IEEE ComSoc.
For more information about the ESD Seminars Series, please contact Karthyek Murthy at firstname.lastname@example.org.