Meta-Learning: Practice and Progress
Zhenguo Li ( Director of AI Theory Lab at Huawei Noah’s Ark Lab )
- 14:00 12th June 2018 ( week Week 8, Trinity Term 2018 )Tony Hoare Room, Robert Hooke Building, entrance next to Museum of Natural History
Abstract:
Meta-learning is a new paradigm in machine learning that aims to learn algorithms (e.g. SGD) instead of models (e.g. CNN). It makes possible a number of previously highly challenging problems such as few-shot learning.
In this talk, I will share some practices in Huawei Noah’s Ark Lab, where we applied meta-learning to a variety of real-world problems in recommender systems, natural language processing, computer vision, and decision making and control.
I will also present some meta-learning algorithms we developed along including Meta-SGD, Federated Meta-Learning, and Deep Meta-Learning.
Bio:
Zhenguo Li is the director of AI Theory Lab at Huawei Noah’s Ark Lab. He received the B.S. and M.S. degrees from the Department of Mathematics at Peking University, and the Ph.D degree from the Department of Information Engineering at the Chinese University of Hong Kong. Before he joined Huawei, he was an associate research scientist in the Department of Electrical Engineering at Columbia University. His research interests include machine learning and artificial intelligence.
Share this:
