BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Enhancing Climate Prediction with Knowledge-Infused Deep Learning 
 Models - Simone Monaco\, Politecnico if Torino
DTSTART:20240926T160000Z
DTEND:20240926T170000Z
UID:TALK222046@talks.cam.ac.uk
CONTACT:Pietro Lio
DESCRIPTION:Modern deep learning models require large amounts of data and 
 computational resources\, often overlooking valuable domain knowledge that
  could enhance their accuracy and efficiency. In this talk\, we present an
  innovative approach that integrates climate science knowledge into deep l
 earning models to boost performance and uncover meaningful relationships b
 etween inputs and outputs. Specifically\, we applied a cutting-edge GraphC
 ast-like architecture for long-term climate prediction\, enhanced by an at
 tention mechanism that accounts for region-specific climate dynamics\, suc
 h as the influence of the El Niño southern oscillator. This enables the m
 odel to capture non-local interactions with greater expressivity. The resu
 lting hybrid model significantly improves prediction accuracy and highligh
 ts key teleconnection patterns\, offering deeper insights into how oscilla
 tors influence global climate states. This method presents a promising ave
 nue for developing more interpretable and precise climate models.
LOCATION:Lecture Theatre 2\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR