Distributed Gradient Tracking for Optimization and Learning over Networks
Keyou You, Associate Professor
Department of Automation, Tsinghua University
Bio: Keyou You received the B.S. degree in Statistical Science from Sun Yat-sen University, Guangzhou, China, in 2007 and the Ph.D. degree in Electrical and Electronic Engineering from Nanyang Technological University (NTU), Singapore, in 2012. After briefly working as a Research Fellow at NTU, he joined Tsinghua University in Beijing, China where he is now an Associate Professor with tenure in the Department of Automation. He held visiting positions at Politecnico di Torino, The Hong Kong University of Science and Technology, The University of Melbourne and etc.
His current research interests include networked control systems, distributed algorithms and learning, and their applications. Dr. You received the Guan Zhaozhi award at the 29th Chinese Control Conference in 2010, and the ACA (Asian Control Association) Temasek Young Educator Award in 2019. He received the National Natural Science Fund for Excellent Young Scholars in 2017.
Abstract: As data gets larger and more decentralized, distributed algorithms over networks provide ample opportunities in many important applications. In this talk, we shall exploit the distributed gradient tracking technique (DGT) to solve large-scale optimization and learning problems, and propose a fully asynchronous DGT which is easy to implement in directed networks with distributed datasets and robust to bounded transmission delays, while maintaining a linear convergence rate if local functions are strongly-convex with Lipschitz-continuous gradients. Moreover, we adopt the DGT to design distributed algorithms with explicit convergence rates for the distributed resource allocation and distributed training over networks, respectively. Experiments are included to show their advantages against the-state-of-the-art algorithms.