Learning edge weights

[edgarsmdn]

Hello!

I am wondering if the following task has been tackle already maybe in other contexts:

Task

I have a graph for which I would like to learn the edge weights. In my dataset, the nodes and connectivity of the graph remain the same, but for different graph-level feature values I have different graph-level output values. From this I would like to learn the edge weights of the graph depending on the given graph-level feature value.

In general, I would be grateful if you can point me to examples of edge weight learning. Thank you!!

[wsad1]

Check out this example which is a link prediction problem. You could simply change the problem to a link regression problem ie. learn edge weights.