Table 2 The performance of ANMDA, ABMDA, GBDT-LR and IRFMDA-100 in 100 times five-fold cross validation
From: ANMDA: anti-noise based computational model for predicting potential miRNA-disease associations
Metrics | ANMDA | ABMDA | GBDT-LR | IRFMDA-100 |
|---|---|---|---|---|
AUROC | 0.9373 ± 0.0005 | 0.9023 ± 0.0021 | 0.9246 ± 0.0010 | 0.9267 ± 0.0009 |
AUPR | 0.9328 ± 0.0008 | 0.8879 ± 0.0032 | 0.9177 ± 0.0015 | 0.9222 ± 0.0012 |
Precision | 0.8561 ± 0.0017 | 0.8213 ± 0.0033 | 0.8403 ± 0.0026 | 0.8447 ± 0.0021 |
Recall | 0.8728 ± 0.0020 | 0.8371 ± 0.0044 | 0.8567 ± 0.0031 | 0.8598 ± 0.0025 |
F1-score | 0.8643 ± 0.0014 | 0.8290 ± 0.0030 | 0.8484 ± 0.0021 | 0.8521 ± 0.0016 |