Skip to main content

Table 1 Results for the active/inactive discrimination task, and Training scheme 1

From: Convolutional architectures for virtual screening

Architecture Bal. accuracy Sensitivity Loss AUC F1-score MCC
Tuned-MLP-Out 0.9880 0.9855 0.0405 0.9979 0.9510 0.9462
Voting 0.9768 0.9710 0.2093 0.9920 0.8965 0.9033
CNN 1D (F) 0.9687 0.9710 0.0688 0.9904 0.8979 0.8813
CNN 2D (R-M-F) 0.9679 0.9565 0.0770 0.9912 0.8918 0.8817
Random Forest (F) 0.9510 0.8985 0.6405 0.9837 0.6065 0.8962
SVM (F) 0.9421 0.8985 0.7883 0.9868 0.8857 0.8731
  1. Fingerprint types: (R)DKIT,(M) organ, (F) eatMorgan, (L)ayered