Table 8 ANOVA for classifier, RUS, hyperparameters and FST as factors of performance in terms of AUPRC
From: IoT information theft prediction using ensemble feature selection
| Â | Df | Sum Sq | Mean Sq | F value | Pr(>F) |
|---|---|---|---|---|---|
Classifier | 7 | 13.34 | 1.91 | 5767.49 | 0.0000 |
RUS | 2 | 0.16 | 0.08 | 249.67 | 0.0000 |
Hyperparameters | 1 | 0.00 | 0.00 | 8.15 | 0.0043 |
FST | 4 | 0.17 | 0.04 | 131.17 | 0.0000 |
Residuals | 11985 | 3.96 | 0.00 | Â | Â |