TABLE 1.
Sensitivity, specificity, and overall accuracy of different machine learning algorithm prediction models for training set and validation set
| Algorithms | Training set (n = 215) | Validation set (n = 55) | ||||
|---|---|---|---|---|---|---|
| Sensitivity | Specificity | Accuracy | Sensitivity | Specificity | Accuracy | |
| KNN | 86.8% | 88.3% | 87.9% | 90.0% | 82.2% | 83.6% |
| SVM | 81.8% | 87.5% | 86.0% | 90.9% | 84.1% | 85.5% |
| XGBoost | 96.0% | 89.7% | 91.2% | 92.3% | 88.1% | 89.1% |
| RF | 98.5% | 99.3% | 99.1% | 93.3% | 92.5% | 92.7% |
| LR | 75.9% | 86.6% | 83.7% | 90.9% | 84.1% | 85.5% |
| DT | 100% | 100% | 100% | 80.0% | 87.5% | 85.5% |
Abbreviations: DT, decision tree; KNN, k‐nearest neighbor; LR, logistic regression; RF, random forest; SVM, support vector machine; XGBoost, extreme gradient boosting.