Speaker: Li Boyang, Nanyang Associate Professor, Nanyang Technological University

Date: Aug. 21, 2021

Time: 16:00-17:00

Location: Tencent Meeting, ID: 682 129 279

Sponsor: School of Software, Shandong University

Abstract:

Deep learning possesses intriguing properties that researchers do not yet fully understand. Deep learning problems usually have many poor local minima but we are usually able to find good optima. Much of this surprising success can be attributed to network initialization and optimization dynamics introduced by stochastic gradient descent. Theoretical understanding and effective utilization of the optimization trajectories thus hold the promise to enhanced performance.

In this talk, I will discuss recent progress on several machine leaming problems that we achieved by utilizing properties of the optimization trajectories. These problems include neural recommendation systems, interpretable deep leaming, and adversarial transfer learning. ming problems that we achieved by utilizing properties of the optimization trajectories. These problems include neural recommendation systems, interpretable deep learning, and adversarial transfer learning.

Bio:

Li Boyang is a Nanyang Associate Professor at the School of Computer Science and Engineering, Nanyang Technological University, Singapore. Prior to NTU, he was a Senior Research Scientist at Baidu Research USA, and a Research Scientist and Group Leader at Disney Research Pittsburgh. He received his Ph.D. degree from Georgia Institute of Technology. In 2021, he was awarded the prestigious NRF Fellowship with research funding of 2.5 million Singapore dollars. His research interests lie mainly in Computational Narrative Intelligence Multi-modal Learning, and Machine Learning.

For more information, please visit:

https://www.view.sdu.edu.cn/info/1020/155570.htm