when label_mat=train_p，Y_value=1，why use (Y_hat_values-1) to calculate the loss when the training data does not distinguish between positive and negative sets of data ?

luoyunan / DTINet

A Network Integration Approach for Drug-Target Interaction Prediction

GNU General Public License v3.0

172 stars 60 forks source link

Open zyyuan180 opened 1 year ago

zyyuan180 commented 1 year ago

luoyunan commented 1 year ago

Could you make it more clear and let us know which lines of code you were referring to? Thanks!