zzaebok/KGCN-pytorch

Why is the result so much better than the original?

Opened this issue · 2 comments

In the case of the music dataset AUC, the original was close to 80%, but this version gets to 85%,

hh-hua commented

+1