Table 4 The performance comparison of different ML models.
From: Non-invasive acoustic classification of adult asthma using an XGBoost model with vocal biomarkers
Model | Accuracy | AUC | Recall | Prec. | F1 | Kappa | MCC | TT (s) |
|---|---|---|---|---|---|---|---|---|
Extreme gradient boosting | 0.8514 ± 0.0622 | 0.9130 ± 0.0774 | 0.8804 ± 0.0758 | 0.8387 ± 0.0757 | 0.8567 ± 0.0594 | 0.7018 ± 0.1254 | 0.7071 ± 0.1275 | 0.015 |
Extra trees classifier | 0.8510 ± 0.0936 | 0.9125 ± 0.0604 | 0.8536 ± 0.1605 | 0.8556 ± 0.0881 | 0.8467 ± 0.1146 | 0.7012 ± 0.1878 | 0.7107 ± 0.1843 | 0.028 |
CatBoost classifier | 0.8505 ± 0.1055 | 0.9286 ± 0.0565 | 0.8500 ± 0.1971 | 0.8498 ± 0.1045 | 0.8378 ± 0.1524 | 0.7009 ± 0.2107 | 0.7130 ± 0.2028 | 1.179 |
Random forest classifier | 0.8443 ± 0.0946 | 0.9157 ± 0.0655 | 0.8268 ± 0.1547 | 0.8645 ± 0.0983 | 0.8374 ± 0.1145 | 0.6882 ± 0.1898 | 0.6982 ± 0.1866 | 0.023 |
Light gradient boosting machine | 0.8371 ± 0.1294 | 0.9010 ± 0.0821 | 0.8268 ± 0.1547 | 0.8481 ± 0.1323 | 0.8345 ± 0.1396 | 0.6730 ± 0.2602 | 0.6762 ± 0.2602 | 0.016 |
Decision tree classifier | 0.8257 ± 0.0675 | 0.8250 ± 0.0662 | 0.8125 ± 0.1036 | 0.8506 ± 0.1066 | 0.8236 ± 0.0663 | 0.6504 ± 0.1336 | 0.6625 ± 0.1385 | 0.004 |
Naive Bayes | 0.7776 ± 0.0773 | 0.8311 ± 0.1033 | 0.7821 ± 0.1143 | 0.7794 ± 0.0816 | 0.7761 ± 0.0817 | 0.5519 ± 0.1557 | 0.5582 ± 0.1552 | 0.004 |
