Replies: 1 comment
-
Hi, If the new nodes represent a different graph (i.e., there are no edges between the training nodes and the test nodes) then you can simply run the prediction on the new graph, ignoring the previous data. Otherwise, you need to feed the full graph to the model. Hope this helps. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi @danielegrattarola,
Given a graph G(V, E), how to do node classification in inductive learning tasks? For example, we contruct a graph for the network training. When new nodes come, how to contruct test dataset and evaluate them?
Thank you
Beta Was this translation helpful? Give feedback.
All reactions