Table 4 Predictive model performance.
From: Metabolomics biomarkers to predict acamprosate treatment response in alcohol-dependent subjects
Performance Index | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Training | Test | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Training | Test | Training | Test | Training | Test | Training | Test | Training | Test | Average | SD | Average | SD | |
True positives | 44 | 17 | 42 | 19 | 43 | 17 | 48 | 22 | 48 | 23 | 45.0 | 2.8 | 19.6 | 2.8 |
True negatives | 13 | 8 | 18 | 6 | 17 | 7 | 7 | 1 | 5 | 3 | 12.0 | 5.8 | 5.0 | 2.9 |
False negatives | 3 | 7 | 5 | 5 | 4 | 7 | 0 | 1 | 0 | 0 | 2.4 | 2.3 | 4.0 | 3.3 |
False positives | 20 | 8 | 15 | 10 | 16 | 9 | 25 | 16 | 27 | 14 | 20.6 | 5.3 | 11.4 | 3.4 |
Sensitivity | 0.936 | 0.708 | 0.893 | 0.792 | 0.915 | 0.708 | 1.000 | 0.957 | 1.000 | 1.000 | 0.949 | 0.049 | 0.833 | 0.138 |
Specificity | 0.394 | 0.500 | 0.545 | 0.375 | 0.515 | 0.438 | 0.219 | 0.059 | 0.156 | 0.176 | 0.366 | 0.174 | 0.310 | 0.186 |
Accuracy | 0.713 | 0.625 | 0.750 | 0.625 | 0.750 | 0.600 | 0.688 | 0.575 | 0.663 | 0.650 | 0.713 | 0.038 | 0.615 | 0.029 |
ROC AUC | 0.832 | 0.659 | 0.845 | 0.578 | 0.802 | 0.570 | 0.756 | 0.724 | 0.771 | 0.703 | 0.801 | 0.038 | 0.647 | 0.071 |
