Can DGL be used for Regression problems too?

Hi,
I was wondering if it is possible to use dgl for regression problems too.
Thank you for the great documentation and tutorials you provided with dgl. I have around 100 graphs which shows connections in human brain (size of each graph is 189*189 and each graph is related to one person’s brain). I also have a continuous variable ranging between 1-30 for each person/graph which shows the person’s performance on a specific task. I want to know if I can predict those behavioral scores based on those brain graphs. So, my problem is a regression problem not classification. Could you let me know how I can use dgl for this? Any comment and help on this is so appreciated!

Thank you!
Samaneh

Yes. For regression problems basically you only need to replace the classification loss with regression loss (say mean squared error).

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.