Table 2. Results of models predicting the success of rest defending situations (successful/unsuccessful rest defense).
Classifier |
Number of features |
Accuracy |
Precision |
f1-Score |
AUC |
Logistic regression (ridge regression regularization) |
17 |
0.87 |
0.56 |
0.57 |
0.76 |
Logistic regression (elastic net regularization) |
17 |
0.82 |
0.54 |
0.54 |
0.74 |
Random Forest Classifier |
17 |
0.84 |
0.55 |
0.56 |
0.78 |
Gradient Boosting |
17 |
0.97 |
0.48 |
0.49 |
0.50 |
XGBoost Classifier |
17 |
0.92 |
0.59 |
0.62 |
0.75 |
AdaBoost Classifier |
17 |
0.96 |
0.63 |
0.61 |
0.60 |
AdaBoost Classifier (excluding distance variables) |
15 |
0.97 |
0.73 |
0.64 |
0.60 |