Table 2.
Results for Burnout vs. No Burnout – Dataset 1 (BNB) (no. test samples = 4071).
| Model | Mean CV Bal. Acc. | Mean CV F1 | Test Bal. Acc. | Test F1 | Test Recall |
|---|---|---|---|---|---|
| Logistic Regression | 0.72 | 0.48 | 0.75 | 0.49 | 0.50 |
| SVM Linear | 0.72 | 0.40 | 0.75 | 0.45 | 0.51 |
| SVM RBF | 0.51 | 0.04 | 0.51 | 0.03 | 0.01 |
| SVM Poly Degree 3 | 0.55 | 0.16 | 0.56 | 0.18 | 0.12 |
| SVM Sigmoid | 0.57 | 0.23 | 0.56 | 0.21 | 0.12 |
| Random Forest | 0.50 | 0.02 | 0.51 | 0.04 | 0.02 |
| Baseline | 0.50 | 0.0 | 0.50 | 0.0 | 0.0 |
Baseline refers to a model predicting no burnout for all samples. The mean CV statistics are computed by taking an average of overall 10 folds cross-validation (CV) during training.