search

SongtaoLiu0823 / LAGNN

[ICML 2022] Local Augmentation for Graph Neural Networks
66 stars 14 forks source link

new issues #4

Closed gaoshihao333 closed 1 year ago

gaoshihao333 commented 1 year ago

cvae_pretrain.py 's U_score = F.nll_loss(output[idx_train], labels[idx_train]) - cross_entropy / args.num_models ,what does it mean?

SongtaoLiu0823 commented 1 year ago

cvae_pretrain.py 's U_score = F.nll_loss(output[idx_train], labels[idx_train]) - cross_entropy / args.num_models ,what does it mean?

It computes the u score based on the Eq.(10) of the paper https://arxiv.org/pdf/2109.03856.pdf