Table 3.
Internal and external validation of AKI prediction models based on different ML algorithms.
| Models | Internal validation |
External validation |
||||||
|---|---|---|---|---|---|---|---|---|
| Accuracy | Recall | PPV | NPV | Accuracy | Recall | PPV | NPV | |
| Bayesian Network | 87.0% | 26.5% | 60.7% | 88.8% | 85.1% | 24.8% | 54.3% | 87.4% |
| Decision tree | 86.1% | 13.5% | 56.2% | 87.2% | 84.6% | 10.9% | 52.4% | 85.7% |
| Random forest | 87.3% | 22.1% | 67.3% | 88.3% | 85.5% | 19.8% | 60.6% | 86.9% |
| Support vector machine | 86.2% | 14.1% | 56.7% | 87.2% | 84.5% | 10.9% | 50.0% | 85.7% |
| Logistic-score | 85.7% | 7.4% | 65.3% | 86.1% | 85.4% | 7.9% | 80.0% | 85.5% |
| Naive Bayes | 86.2% | 30.5% | 53.3% | 89.2% | 84.9% | 28.7% | 52.7% | 87.9% |
Accuracy rate is the sum of correctly classified cases test divided by the data set size (TP + TN)/(TP + TN + FP + FN). Recall rate is the positively classified cases divided by the positive cases TP/(TP + FN). Positive predictive value (PPV) is the proportion of positive cases that are true positives TP/(TP + FP). Negative predictive value (NPV) is the proportion of negative cases that are true negatives TN/(TN + FN).