Table 8.
Validation sample results for AF detection
| FSM | AUC | Sensitivity | Specificity | NSF | Runtime (s) |
|---|---|---|---|---|---|
| FULL | 95.48 | 91.26 | 99.7 | 16 | 0.00 |
| CFS | 99.51 | 98.61 | 96.33 | 3 | 0.11 |
| CHI2 | 95.28 | 90.90 | 99.67 | 15 | 0.83 |
| GAMMA_BACK | 98.74 | 95.08 | 97.40 | 7 | 2.12 |
| GAMMA_BF | 98.74 | 95.08 | 97.40 | 7 | 0.91 |
| GAMMA_FORW | 98.74 | 95.08 | 97.40 | 7 | 0.80 |
| LASSO | 91.27 | 82.87 | 99.66 | 7 | 20.84 |
| RFI | 66.22 | 56.52 | 79.36 | 1 | 178.29 |
| STEP | 98.27 | 94.75 | 98.07 | 6 | 52.84 |
| SU | 95.28 | 90.90 | 99.67 | 15 | 0.87 |
| SVM-RFE | 93.46 | 87.17 | 99.75 | 14 | 43.97 |
The best value of each indicator is highlighted with bold and underline. The second-best value is highlighted in bold. The indicators for the FULL model are not highlighted in the table since the focus of the study is to compare the feature selection methods
Abbreviation: NSF Number of Selected Features