Thomas Kipf

Even if you perform transductive learning with a GCN model the model will be able to predict in an inductive setting. The model generally has a favorable inductive bias that...

Yes! On Mon 8. Apr 2019 at 11:20 Ye D. wrote: > Hi, > If I trained a gcn model,can I use it to generate embeddings to unseen > nodes,which...

The adjacency matrix you provide during training can be different from the adjacency matrix provided at test time. You can make a prediction for a new data point by respectively...

Dimensions need to match if you do not re-train the model. On Mon, May 6, 2019 at 1:39 PM crystal0913 wrote: > Does the dimension match of trained weight W...

Hi, Thanks for your question. 1) Here's a paper that summarizes this idea nicely: https://arxiv.org/abs/1704.06199 2) Yes, this is correct.

The main novelty of GraphSAGE is a neighborhood sampling step (but this is independent of whether these models are used inductively or transductively). You can think of GraphSAGE as GCN...

No worries :) I'll leave this open since this question seems to appear quite frequently (so others can more easily find it).

Thanks, I’ll leave this open for reference. On Sat 4. May 2019 at 00:10 XinliYu wrote: > The two "print" lines in the following code checks the outputs are the...

In terms of loading the .x/.tx files, maybe try python 2.7. In terms of loading your own data: you don't need to provide .x/.tx files, but instead you can provide...

This is strange. Maybe try the data loader in https://github.com/tkipf/pygcn -- note, however, that the dataset splits here are different. On Sat, May 11, 2019 at 10:39 AM aubutt wrote:...