Skip to main content

Table 1 The optimal parameters of DPDDI

From: DPDDI: a deep predictor for drug-drug interactions

Parameters

L-rate

Epochs

Dropout

B-size

I-dima

H-dim

O-dimb

Feature extractor

0.001

1400

0.0001

Full-batch

1562

[512,128]

128

Predictor

0.01

140

0.001

50

256

[128,64,32]

2

  1. aI-dim denotes the neuro numbers in input layer; bO-dim denotes the neuro numbers in output layer