Table 4.
Threshold discriminative metric evaluation for show and no-show appointment prediction.
| Algorithm | Accuracy | Precision | Recall | F1-score |
|---|---|---|---|---|
| Random forest | 84.96% | 80% | 93% | 86.25% |
| Decision tree | 85.18% | 82% | 90% | 86.42% |
| Logistic regression | 85.35% | 83% | 89% | 86.18% |
| XG Boost | 76.69% | 80% | 83% | 81.44% |
| Gradient boosting | 73.53% | 76% | 77% | 77.32% |
| Adaboost | 70.46% | 74% | 72% | 73.17% |
| SVM | 67.09% | 69% | 74% | 72.15% |
| Naive Bayes | 63.98% | 66% | 70% | 68.44% |
| SGD | 67.14% | 60% | 84% | 70.18% |
| Multilayer perceptron | 80.93% | 64% | 77% | 70.52% |