From: Machine Learning for Understanding and Predicting Injuries in Football
 | Machine learning algorithms | Pre-processing techniques | Accuracy (%) | Precision (%) | AUC | Recall (%) | F1-score (%) | Specificity (%) |
---|---|---|---|---|---|---|---|---|
[33] | Decision tree | Feature selection, oversampling–SMOTE | – | 50 | 0.76 | 80 | 64 | – |
[35] | SVM | Data normalisation | 83.50 | – | – | – | – | – |
[36] | SmoteBoost | Oversampling—SMOTE | – | – | 0.75 | 65.90 | – | 79.10 |
[37] | SmoteBoost | Oversampling—SMOTE | – | – | 0.84 | 77.80 | – | 83.80 |
[39] (a) | XGBoost | Unmentioned | – | 85 | – | 85 | 85 | – |
[39] (b) |  | – | 78 | – | 78 | 78 | – | |
[41] | Decision tree | Various balancing techniques | – | – | 0.66 | 55.60 | – | 74.20 |
[42] (a)* | Random forest | Missing values imputation | 95.5 | 92.2 | 0.92 | 94.5 | – | – |
[42] (b)* | XGBoost | 97 | 97 | 0.97 | 97 | – | – | |
[43] | Cox regression | Unmentioned | – | – | – | – | – | – |
[44] (a) | Supervised principal components analysis | Unmentioned | 88.80 | 55 | – | 33 | – | – |
[44] (b) |  | 97.07 | 19 | – | 20 | – | – |