January 20, 2023
ECE Research Seminar
Friday, January 27 12 – 1:00 p.m.
Exploiting Networked Intelligence Using Federated Learning
Chun-Hung Liu | email@example.com
Abstract: The thriving of machine learning (ML) technologies is driving the evolution of solving many diverse scientific and engineering problems from model-based approaches to data-driven ones. How to effectively exploit a considerable amount of data distributed over a large territory safely and efficiently is the key to making ML have much more transformative impacts on data-driven solutions to complex real-world problems. Conventional ML conducted by a server in a centralized fashion requires raw data transmitted from clients to the server over networks, and thereby it causes many practical issues, such as privacy, security, network congestion, etc. As a result, it cannot efficiently exploit “networked intelligence” by learning from big data distributed over a huge network. Federated learning (FL) recently proposed can make the server conduct a learning task without raw data delivered from its clients, and it accordingly is a viable approach to networked intelligence. Nonetheless, the performances of FL are fairly sensitive to heterogeneous network conditions, such as data and client heterogeneity, network resource, networking condition, etc. This talk will introduce a new FL model that is resilient to network variations and untethered by practical networking and communication limitations. The new resilient FL model aims to achieve high learning performances in terms of reliability, scalability, and adaptability. As such, it is expected to significantly improve the fundamental performance limits of FL with system impairment and data heterogeneity.
Dr. Chun-Hung Liu is currently an assistant professor in the Department of Electrical and Computer Engineering at Mississippi State University (MSU). Before joining MSU, he was with the University of Michigan at Dearborn and National (Yang Ming) Chiao Tung University in Taiwan. He received a B.S. degree from National Taiwan University, Taipei, Taiwan, an M.S. degree from Massachusetts Institute of Technology, and a Ph.D. degree in Electrical and Computer Engineering from the University of Texas at Austin. He was a recipient of the Best Paper Award from the IEEE Globecom Conference in 2008 and 2014, and a recipient of the Young Research Scholar Award from the Ministry of Science and Technology of Taiwan in 2015. His research interests lie in wireless networking, machine learning, data science, stochastic control, optimization theory, etc.
For further information, contact: Dr. Jenny Du | firstname.lastname@example.org | 662-325-2035
For WebEx information, scan the QR Code