BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Weisfeiler and Lehman Go Cellular: CW Networks - Cristian Bodnar (
 University of Cambridge) and Fabrizio Frasca (Imperial College\, Twitter)
DTSTART:20211116T131500Z
DTEND:20211116T141500Z
UID:TALK165856@talks.cam.ac.uk
CONTACT:Mateja Jamnik
DESCRIPTION:"Join us on Zoom":https://zoom.us/j/99166955895?pwd=SzI0M3pMVE
 kvNmw3Q0dqNDVRalZvdz09\n\nGraph Neural Networks (GNNs) are limited in thei
 r expressive power\, struggle with long-range interactions and lack a prin
 cipled way to model higher-order structures. These problems can be attribu
 ted to the strong coupling between the computational graph and the input g
 raph structure. The recently proposed Message Passing Simplicial Networks 
 naturally decouple these elements by performing message passing on the cli
 que complex of the graph. Nevertheless\, these models can be severely cons
 trained by the rigid combinatorial structure of Simplicial Complexes (SCs)
 . In this work\, we extend recent theoretical results on SCs to regular Ce
 ll Complexes\, topological objects that flexibly subsume SCs and graphs. W
 e show that this generalisation provides a powerful set of graph "lifting"
  transformations\, each leading to a unique hierarchical message passing p
 rocedure. The resulting methods\, which we collectively call CW Networks (
 CWNs)\, are strictly more powerful than the WL test and not less powerful 
 than the 3-WL test. In particular\, we demonstrate the effectiveness of on
 e such scheme\, based on rings\, when applied to molecular graph problems.
  The proposed architecture benefits from provably larger expressivity than
  commonly used GNNs\, principled modelling of higher-order signals and fro
 m compressing the distances between nodes. We demonstrate that our model a
 chieves state-of-the-art results on a variety of molecular datasets.
LOCATION:Zoom
END:VEVENT
END:VCALENDAR