BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250626T233532Z
LOCATION:B302-B305
DTSTART;TZID=America/New_York:20241121T100000
DTEND;TZID=America/New_York:20241121T170000
UID:submissions.supercomputing.org_SC24_sess534_post130@linklings.com
SUMMARY:A Novel Gradient Compression Design with Ultra-High Compression Ra
 tio for Communication-Efficient Federated Learning
DESCRIPTION:Zhijing Ye and Xiaodong Yu (Stevens Institute of Technology)\n
 \nFederated learning is a privacy-preserving machine learning approach. It
  allows numerous geographically distributed clients to collaboratively tra
 in a large model while maintaining local data privacy. In heterogeneous de
 vice settings, limited network bandwidth is a major bottleneck that constr
 ains system performance. In this work, we propose a novel gradient compres
 sion method for federated learning that aims to achieve communication effi
 ciency and a low error floor by estimating the prototype of gradients on b
 oth the server and client sides and sending only the difference between th
 e real gradient and the estimated prototype. This approach further reduces
  the total bits required for model updates. Additionally, the memory requi
 rement will be lighter on the client side but heavier on the server side c
 ompared to traditional error feedback methods. Experiments on training neu
 ral networks show that our method is more communication-efficient with lit
 tle impact on training and test accuracy.\n\nRegistration Category: Tech P
 rogram Reg Pass, Exhibits Reg Pass\n\nSession Chairs: Ayesha Afzal (Friedr
 ich-Alexander University, Erlangen-Nuremberg; Erlangen National High Perfo
 rmance Computing Center); Sally Ellingson (University of Kentucky); and Al
 an Sussman (University of Maryland)\n\n
END:VEVENT
END:VCALENDAR