BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//scipy-2026//talk//ECYVWR
BEGIN:VTIMEZONE
TZID:CST
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20061029T080000Z
TZNAME:CST
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
END:STANDARD
BEGIN:STANDARD
DTSTART:20071104T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:CST
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000402T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060402T090000Z
TZNAME:CDT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20070311T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:CDT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-scipy-2026-ECYVWR@pretalx.com
DTSTART;TZID=CST:20260716T135500
DTEND;TZID=CST:20260716T142500
DESCRIPTION:Deploying deep learning models for time-series forecasting at r
 etail scale presents a fundamental tension between prediction accuracy and
  computational cost. This talk presents a Python-based framework combining
  structured pruning\, quantization-aware training\, and knowledge distilla
 tion to compress LSTM networks for demand forecasting. Using NumPy\, Tenso
 rFlow/Keras\, and scikit-learn\, we achieved 47% accuracy improvement over
  baseline models while reducing model size by 73% and inference costs by 9
 2%. We discuss practical implementation patterns\, reproducibility conside
 rations\, and how these compression techniques generalize beyond retail to
  any domain requiring efficient sequential prediction at scale.
DTSTAMP:20260622T110107Z
LOCATION:Memorial Hall
SUMMARY:Compressing LSTM Networks for Scalable Retail Demand Forecasting: A
  Python-Based Approach to Efficient Time-Series Prediction - Ravi Teja Pag
 idoju
URL:https://pretalx.com/scipy-2026/talk/ECYVWR/
END:VEVENT
END:VCALENDAR