Table 3.
Mortality prediction performance of selected classifiers on various reduced data sets extracted via chi-square significance.
| Model and data set granularity | Average values across folds (%) | |||||||
|
|
Specificitya (95% CI) | Sensitivitya (95% CI) | Accuracya (95% CI) | AUCb (95% CI) | ||||
| Random forest | ||||||||
|
|
25-feature data set | 83.2 (80.1-86.3) | 89.1 (86.8-91.4) | 89.2 (88.0-90.4) | 96.3 (95.9-96.6) | |||
|
|
7-feature data set | 90.7 (89.1-92.3) | 92.0 (91.0-92.9) | 92.5 (91.9-93.0) | 95.5 (95.2-95.8) | |||
|
|
1-feature data set | 84.8 (83.3-86.2) | 77.6 (76.6-78.6) | 81.1 (80.7-81.6) | 88.5 (88.1-89.0) | |||
| Logistic regression | ||||||||
|
|
25-feature data set | 82.9 (79.9-85.9) | 79.6 (76.0-83.2) | 83.5 (81.9-85.1) | 88.6 (87.5-89.7) | |||
|
|
7-feature data set | 90.4 (89.1-91.8) | 69.9 (64.9-74.8) | 79.8 (77.3-82.1) | 89.5 (88.4-90.6) | |||
|
|
1-feature data set | 80.2 (79.4-81.0) | 80.9 (79.8-82.1) | 81.6 (81.1-82.1) | 84.2 (83.5-84.9) | |||
aReported performance metrics represent averages across multiple simulations of 3-fold cross validation and, due to class balance variation between folds, accuracy metrics are not always a weighted average of their sensitivity and specificity.
bAUC: area under the curve obtained from the receiver operating characteristic curve.