Technical Program

WE2-4: Distributed Coding and Optimization

Session Type: Virtual
Time: Wednesday, November 4, 13:00 - 14:20
Virtual Session: View on Virtual Platform
Location: Room 4
Session Chair: Gauri Joshi, Carnegie Mellon University
 
WE2-4-1: Robust Class Parallelism - Error Resilient Parallel Inference with Low Communication Cost
         Yaoqing Yang; UC Berkeley
         Jichan Chung; UC Berkeley
         Guanhua Wang; UC Berkeley
         Vipul Gupta; UC Berkeley
         Adarsh Karnati; UC Berkeley
         Kenan Jiang; UC Berkeley
         Ion Stoica; UC Berkeley
         Joseph Gonzalez; UC Berkeley
         Kannan Ramchandran; UC Berkeley
 
WE2-4-2: vqSGD: Vector Quantized Stochastic Gradient Descent
         Venkata Gandikota; University of Massachusetts Amherst
         Daniel Kane; University of California San Diego
         Raj Kumar Maity; University of Massachusetts Amherst
         Arya Mazumdar; University of Massachusetts Amherst
 
WE2-4-3: Bandit-based Communication-Efficient Client Selection Strategies for Federated Learning
         Yae Jee Cho; Carnegie Mellon University
         Samarth Gupta; Carnegie Mellon University
         Gauri Joshi; Carnegie Mellon University
         Osman Yağan; Carnegie Mellon University
 
WE2-4-4: Attack of the Tails: Yes, You Really Can Backdoor Federated Learning
         Hongyi Wang; University of Wisconsin-Madison
         Kartik Sreenivasan; University of Wisconsin-Madison
         Harit Vishwakarma; University of Wisconsin-Madison
         Jy-yong Sohn; KAIST
         Shashank Rajput; University of Wisconsin-Madison
         Saurabh Agarwal; University of Wisconsin-Madison
         Kangwook Lee; University of Wisconsin-Madison
         Dimitris Papailiopoulos; University of Wisconsin-Madison