Tensor Learning Team Seminar (Talk by Dr. Quanming Yao, Tsinghua University).

Title: Automated Learning from Knowledge Graph

Speaker: Dr. Quanming Yao, Tsinghua University, China.

Abstract: Knowledge graph (KG) is a kind of graph-structured data containing symbols and concepts, and the recent interests of machine learning have moved towards learning from knowledge enhanced and complex structure data. Thus, learning from KG has become an important topic with lots of research interests. In this presentation, we will talk about our recent works on automated KG learning. First, we will show how to search data-specific bi-linear models which can better explore semantics in KG than existing hand-designed one. Then, we look inside hyperparameters in bi-linear models and show how these hyperparameters can be efficiently tuned. Related works are recently published in IEEE TPAMI, NeurIPS, ACL, KDD and etc. The proposed method also holds the 1st place in relevant KG task on the well-known open graph benchmark.

Bio: Dr. Quanming Yao currently is a tenure-track assistant professor at Department of Electronic Engineering, Tsinghua University. Before that, he spent three years from a researcher to a senior scientist in 4Paradigm INC, where he set up and led the company's machine learning research team. He obtained his Ph.D. degree at the Department of Computer Science and Engineering of Hong Kong University of Science and Technology (HKUST) in and received his bachelor degree at HuaZhong University of Science and Technology (HUST). He regularly serves as area chairs of ICML, NeurIPS, ICLR and ACML. He is also a receipt of National Youth Talent Plan (China), Forbes 30 Under 30 (China), Young Scientist Awards (Hong Kong Institution of Science), and Google Fellowship (in machine learning).