Table 7.
Classification performance for models on Datasets B and C.
| Dataset B test performance (mean (and std. dev.) across 10 tests; training on 90%, testing on 10%) Actual -in rates (mean and std.dev): 62.4% (5.2) |
Dataset C performance (training on 100% Dataset B, testing on Dataset C) Actual -in rates: 60.3% |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pred. -in rate | Accuracy | Precision | Recall | F1 score | AUC | Pred. -in rate | Accuracy | Precision | Recall | F1 score | AUC | |
| Linear SVM | 64.5% (5.3) | 78.2% (5.6) | 0.84 (0.05) | 0.81 (0.06) | 3.25 (0.23) | 0.76 (0.06) | 70.7% | 78.7% | 0.91 | 0.78 | 3.10 | 0.75 |
| Radial SVM | 66.6% (4.8) | 86.1% (2.8) | 0.92 (0.04) | 0.86 (0.05) | 3.46 (0.20) | 0.84 (0.04) | 69.0% | 79.3% | 0.90 | 0.79 | 3.15 | 0.76 |
| RF | 69.9% (5.6) | 82.1% (3.9) | 0.92 (0.03) | 0.82 (0.06) | 3.28 (0.24) | 0.79 (0.05) | 72.5% | 77.7% | 0.92 | 0.76 | 3.05 | 0.74 |