n graphs，how to train

zxj32 / GCN-GRU

Deep Learning for Predicting Dynamic Uncertain Opinions in Network Data

MIT License

25 stars 4 forks source link

n graphs，how to train #3

Open maybewh opened 4 years ago

maybewh commented 4 years ago

If I have n graphs which every node has a label， how I can get spatial feature and train in time series？

zxj32 commented 4 years ago

So you don't have the physical feature for each node, right? Then we can use the identity matrix as the spatial feature for input (line 49 in "GCN_GRU_run.py")