Table 2.
AUC comparison of six prediction models
| Prediction Model | AUC | SE | 95% CI |
|---|---|---|---|
| Logistic | 0.854 | 0.0227 | 0.806 to 0.894 |
| Lasso | 0.855 | 0.0224 | 0.808 to 0.895 |
| LDA | 0.853 | 0.0228 | 0.806 to 0.893 |
| Naïve Bayes1 | 0.840 | 0.0239 | 0.791 to 0.881 |
| SVM1 | 0.855 | 0.0226 | 0.808 to 0.895 |
| Random Forest | 0.842 | 0.0237 | 0.793 to 0.884 |
Among the six models evaluated, the multivariable logistic regression model exhibited no statistically significant distinction in comparison to other models (pairwise comparison P > 0.05)
AUC area under the ROC curve, LDA linear discriminant analysis, SVM support vector machine
1The difference between AUC of Naïve Bayes and SVM was statistically significant