Qiang Yang
Fellow of the Canadian Academy of Engineering and the Royal College of Canada
Qiang Yang is a Fellow of the Engineering Institute of Canada and the Royal Institution of Canada, Chief Artificial Intelligence Advisor of Microfinance, Professor Emeritus of the Hong Kong University of Science and Technology, Chairman of AAAI-2021, and Former Chairman of the Board of Governors of the International Joint Committee on Artificial Intelligence (IJCAI). Donald E. Walker Distinguished Service Award for 2023. He is also the founding editor-in-chief of two top international journals, ACM TIST and IEEE TRANS on BIG DATA, and a Fellow of several international societies, including CAAI/AAAI/ACM/IEEE/AAAS, etc. His research interests include transfer learning and federated learning research and applications, and he has published books such as Transfer Learning, Federated Learning, Privacy in Computing, and Federated Learning in Action. Computing” and ‘Federated Learning in Practice’.
Topic
Federated Size Model Collaborative Learning
The development of large models encounters data and arithmetic bottlenecks, and federated learning provides a new avenue for AI. We believe that the future of AI is the collaboration of large and small models, allowing large models in the cloud and localized small models to be augmented together through federated learning and transfer learning. This protects the privacy and data security of all parties while the models improve their capabilities. In this talk, I will discuss how the Federated Big Model framework can be used for teacher-student collaborative learning tasks in the context of big models. I will first review the development of AI and the concept of federated learning, and then discuss how federated learning, transfer learning, and big models can be organically combined to make the continued development and application of big models smoother and more efficient.
