Enabling Effective and Efficient Federated Learning at Future Network Edge

Add to your list(s) Download to your calendar using vCal

• Dr. Bing Luo, Duke Kunshan University
• Monday 11 March 2024, 12:00-13:00
• Room 909b, EEE Department.

If you have a question about this talk, please contact Kin K Leung.

In the era of advancing IoT and social network applications, the surge of data generated by billions of smart devices at the network edge holds immense potential for intelligent applications. However, the conventional approach of centralizing this data in data centers or the cloud raises serious data privacy concerns. To address this challenge, Federated Learning (FL) has emerged as an attractive distributed AI paradigm, enabling network-edge clients to collaboratively train machine learning models while keeping their raw data private and decentralized.

In this talk, I will first provide a brief overview of FL’s unique characteristics and technical challenges, such as system and statistical heterogeneity. Then, I will highlight some of our research efforts focused on designing effective and efficient FL at the resource-constrained network edge. This includes optimization-based approach featuring adaptive parameter control and importance sampling, as well as a game-theoretical unbiased incentive mechanism. Finally, I will introduce our recently launched project, FedCampus, which applies FL and differential privacy techniques to learn population health insights at Duke Kunshan University.

About the Speaker: Dr. Bing Luo is an Assistant Professor of Data and Computational Science at Duke Kunshan University (DKU). Prior to joining DKU , he was a joint Postdoc Researcher at The Chinese University of Hong Kong (Shenzhen) and Yale University. He received his Ph.D. from The University of Melbourne, and the B.E. and M.S. degrees from Beijing University of Posts and Telecommunications (BUPT). Before pursuing his Ph.D, he gained several years of industry experience as a project manager at China Mobile Corporation Headquarter. His current research interests include federated learning and analytics, embedded AI for IoT and mobile systems. For more information, please visit his webpage: https://luobing1008.github.io/

This talk is part of the Featured talks series.

This talk is included in these lists:

• Andrea Picciau's list
• COMMSP & CP list
• COMMSP Seminar
• Featured talks
• Room 909b, EEE Department
• SAP Talks

Note that ex-directory lists are not shown.