K-plex cover pooling for graph neural networks

HIGHLIGHTS

• who: Davide Bacciu from the Department of Computer Science, Università di Pisa, Pisa, Italy have published the research work: K-plex cover pooling for graph neural networks, in the Journal: (JOURNAL)
• what: The authors show how KPlexPool, despite being nonadaptive, provides a flexible and robust definition for node communities which can generalize well to graphs of different nature and topology, from molecular structures to social networks. The authors propose KPlexPool, the first pooling mechanism based on the concepts of k-plex communities and graph covers (Sect 2.3). - The authors propose a post-processing cover . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.