Asynchronous optimization and learning with delay-free step-sizes



    Department of Systems Engineering and Engineering Management

                       The Chinese University of Hong Kong


Date: Thursday, January 25, 4:30 pm – 5:30 pm

Venue: ERB 513, The Chinese University of Hong Kong

Title:Asynchronous optimization and learning with delay-free step-sizes

Speaker: Dr. Xuyang Wu



In networked systems, many tasks such as federated learning, economic dispatch in power systems, and multi-robot coordination can be formulated as distributed optimization or learning problems. When solving these problems, asynchronous algorithms often have enhanced efficiency, implementation flexibility, and robustness against single-node failures compared to their synchronous counterparts. However, existing asynchronous algorithms often use an upper bound on the information delays in the system to determine step-sizes. Not only are the delay bounds hard to obtain in advance, but they also result in unnecessarily small step-sizes and slow convergence.In this talk, I will share our recent efforts in addressing this issue. Unlike most existing works, our proposed algorithms can converge with step-size conditions that do not rely on any delay information, but still converge to the same fixed-point set of their synchronous counterparts. This feature brings benefits in both theory and practical performance.


Xuyang Wu received the B.S degree in Applied Mathematics from Northwestern Polytechnical University, China, in 2015, and the Ph.D. degree in Communication and Information Systems from the University of Chinese Academy of Sciences, China, in 2020. From 2020 - 2023, He was a postdoctoral researcher at KTH Royal Institute of Technology, Sweden. He is currently visiting Professor Hoi-To Wai in the Department of Systems Engineering and Engineering Management, CUHK. His research interests include distributed and large-scale optimization, machine learning, and related areas. He has published several first-authored papers in top-tier journals in the control society and machine learning conferences.

Everyone is welcome to attend the talk!

SEEM-5202 Website:


Thursday, January 25, 2024 - 16:30