BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260402T024533Z
LOCATION:3001\, Level 3
DTSTART;TZID=America/Los_Angeles:20250625T103000
DTEND;TZID=America/Los_Angeles:20250625T104500
UID:dac_DAC 2025_sess114_RESEARCH909@linklings.com
SUMMARY:PracMHBench: Re-evaluating Model-Heterogeneous Federated Learning 
 Based on Practical Edge Device Constraints
DESCRIPTION:Yuanchun Guo, Bingyan Liu, yulong sha, and zhensheng xian (Bei
 jing University of Posts and Telecommunications)\n\nFederating heterogeneo
 us models on edge devices with diverse resource constraints has been a not
 able trend in recent years. Compared to traditional federated learning (FL
 ) that assumes an identical model architecture to cooperate, model-heterog
 eneous FL is more practical and flexible since the model can be customized
  to satisfy the deployment requirement. Unfortunately, no prior work ever 
 dives into the existing model-heterogeneous FL algorithms under the practi
 cal edge device constraints and provides quantitative analysis on various 
 data scenarios and metrics, which motivates us to rethink and re-evaluate 
 this paradigm. In our work, we construct the first system platform \textbf
 {PracMHBench} to evaluate model-heterogeneous FL on practical constraints 
 of edge devices, where diverse model heterogeneity algorithms are classifi
 ed and tested on multiple data tasks and metrics. Based on the platform, w
 e perform extensive experiments on these algorithms under the different ed
 ge constraints to observe their applicability and the corresponding hetero
 geneity pattern.\n\nTopics: AI\n\nTracks: AI4: AI/ML System and Platform D
 esign\n\nSession Chairs: Xiaoxuan Yang (University of Virginia, Stanford U
 niversity) and Shihao Song (Nvidia)\n\n
END:VEVENT
END:VCALENDAR