RE-GCN is comprised of four steps: 1) selecting initial structure centers; 2) refining the structure centers; 3) expanding pseudo-labeled set; and 4) finally training GCN for community partition.

RE-GCN is comprised of four steps: 1) selecting initial structure centers; 2) refining the structure centers; 3) expanding pseudo-labeled set; and 4) finally training GCN for community partition.

<p>RE-GCN is comprised of four steps: 1) selecting initial structure centers; 2) refining the structure centers; 3) expanding pseudo-labeled set; and 4) finally training GCN for community partition.</p>

Saved in:

Bibliographic Details
Main Author:	Bing Guo (25387) (author)
Other Authors:	Liping Deng (756374) (author), Tao Lian (10662498) (author)
Published:	2025
Subjects:	Biochemistry Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified refinement process yields extensive experiments demonstrate corresponding induced subgraph structure centers based representative structure centers refined structure centers initial structure centers inappropriate structure centers highest structure importance current structure centers label propagation ability temporary graph partition structure center refinement realize community detection community detection performance graph convolutional network labeled set expansion gcn ), considering structure center community structure network topology label information various graph labeled set labeled nodes graph topology graph convolution entire graph two steps structured data shallow gcn recently exploited proposed approach localized filter limited amount iteratively refine highly dependent gcn improves final gcn efficient approach classical problem adverse effect
Tags:	Add Tag No Tags, Be the first to tag this record!

Cannot write session to /tmp/vufind_sessions/sess_micc37s4c13vs4vfksdelmes9o