Table 6.
Performances of the six machine learning classifiers for predicting AR expression.
| Classifier | AUC (95% CI) | ACC (%) | SEN (%) | SPE (%) | NPV (%) | PPV (%) |
|---|---|---|---|---|---|---|
| MLP | 0.907 (0.851–0.947) | 85.8 | 85.6 | 86.7 | 57.8 | 96.6 |
| LDA | 0.880 (0.820–0.926) | 81.5 | 79.6 | 90.0 | 50.0 | 97.2 |
| SVM | 0.852 (0.788–0.903) | 85.8 | 87.9 | 76.7 | 59.0 | 94.3 |
| GNB | 0.881 (0.821–0.927) | 80.9 | 80.3 | 83.3 | 49.0 | 95.5 |
| RF | 0.905 (0.849–0.945) | 82.1 | 79.6 | 93.3 | 50.9 | 98.1 |
| LR | 0.888 (0.829–0.932) | 86.4 | 88.6 | 76.7 | 60.5 | 94.4 |
SVM, Support Vector Machine (radial bias function); RF, Random Forest; LR, Logistic Regression; MLP, Multilayer Perceptron; GNB, Gaussian Naïve Bayes; LDA, Linear Discriminant Analysis; AUC, the area under curve; ACC, accuracy; SEN, sensitivity; SPE, specificity; NPV, negative predictive value; PPV, positive predictive value; AR, androgen receptor.