BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Problem of Size Generalization in Graph Neural Networks - Davi
 de Buffelli \, Universita' di Padova
DTSTART:20220704T160000Z
DTEND:20220704T170000Z
UID:TALK176399@talks.cam.ac.uk
CONTACT:Pietro Lio
DESCRIPTION: In the past few years\, graph neural networks (GNNs) have bec
 ome the de facto model of choice for graph classification and other tasks 
 on graph structured data. While\, from the theoretical viewpoint\, most GN
 Ns can operate on graphs of any size\, it is empirically observed that the
 ir classification performance degrades when they are applied on graphs wit
 h sizes that differ from those in the training data. In this talk we will 
 give an overview of the current approaches to tackle the issue of poor siz
 e-generalization in GNNs\, and we will introduce our recent work in this a
 rea.
LOCATION:Lecture theatre and zoom (https://cl-cam-ac-uk.zoom.us/j/96155557
 207?pwd=VXlyUVZidVRxWFRaWS9tak1SUjV3Zz09  ID: 961 5555 7207 Passcode: 9137
 48)
END:VEVENT
END:VCALENDAR