Skip to main content
. 2022 Jun 2;38(14):3582–3589. doi: 10.1093/bioinformatics/btac377

Table 1.

Hyperparameter setting in MINN-DTI

Hyperparameter Value
Learning rate 0.0001
Number of residual blocks 32
Number of filters 32
Graph feature size 32
Attention heads 8
Hidden size of Decoder 32
Iterations of message passing 4
Dropout 0.2