Table 5 Performance on blind dataset using best deep neural network classifier
From: Identification of infectious disease-associated host genes using machine learning techniques
Best Model Features set | Vector length | P(+): N(−) | Sensitivity (%) | Specificity (%) | Accuracy (%) | PPV (%) | MCC | F1 score (%) | AUC |
|---|---|---|---|---|---|---|---|---|---|
PAAC_Network properties | 59 | 1: 1 | 85.09 | 76.32 | 80.70 | 78.23 | 0.62 | 81.51 | 0.872 |
Selected Features For PAAC _Network properties | 16 | 1: 1 | 89.47 | 77.19 | 83.33 | 79.69 | 0.67 | 84.30 | 0.904 |
Selected Features For Normalized And Filtered PAAC_ Network properties | 10 | 1: 1 | 88.60 | 80.70 | 84.65 | 82.11 | 0.70 | 85.23 | 0.879 |