Date/Time
Date(s) - Mon 26 June
13:30 - 14:30
Location
Cotton Club, Cotton 350, VUW
CDSAI, ECS and IEEE CIS/NZ Central Section Seminar
Speaker: Prof Haibo He (IEEE Fellow, University of Rhode Island)
Abstract
Abstract: The recently advancements in artificial/computational intelligence (AI/CI) has witnessed tremendous excitements worldwide from academia and industry, which demonstrated the power of AI/CI over complicated tasks. This talk aims to review and discuss the recent research developments in computational intelligence with a focus on the key characteristics of reinforcement learning (RL). Specifically, I will present a new reinforcement learning and adaptive dynamic programing (RL/ADP) framework for improved decision-making capability. This framework integrates a hierarchical goal generator network to provide a more informative and detailed internal goal representation to guide the decision-making process. Compared to the existing RL/ADP approaches with a manual or “hand-crafted” reinforcement signal design, this framework can automatically and adaptively develop the internal goal representation over time, without the requirement of an explicit external reward. Detailed learning architecture and associated learning algorithms will be discussed in this seminar.
Brief Bio: Haibo He is the Robert Haas Endowed Chair Professor of Electrical Engineering at the University of Rhode Island, USA. He has published one sole-author book (Wiley), edited one book (Wiley-IEEE) and six conference proceedings (Springer), and authored/co-authored numerous peer-reviewed journal and conference papers, including several highly cited papers, Transaction cover page highlighted paper, and Best Paper awards. He has been a long-time volunteer for the IEEE Computational Intelligence Society (IEEE CIS) and served/serves at various capacities, ranging from the Chair of the IEEE CIS Neural Network Technical Committee (NNTC), Vice Chair of the IEEE CIS Fellow Committee, General Chair of the IEEE Symposium Series on Computational Intelligence (IEEE SSCI 2014), Plenary Co-Chair of the IJCNN 2023, to the Editor-in-Chief (EIC) of the IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS, 2016 – 2021), a flagship journal of the IEEE CIS. He was a recipient of the IEEE CIS Outstanding Early Career Award (2014) and National Science Foundation CAREER Award (2011).