Table 3.
Model | SVM | Random forest | Multiple logistic regression | ||||||
---|---|---|---|---|---|---|---|---|---|
Index | Accuracy | AUROCa | P valueb | Accuracy | AUROC | P value | Accuracy | AUROC | P value |
Men ≥ 65 years | 0.853 | 0.822 | .003 | 0.846 | 0.837 | .002 | 0.843 | 0.805 | .002 |
Women ≥ 65 years | 0.838 | 0.835 | .015 | 0.826 | 0.819 | .001 | 0.826 | 0.796 | .003 |
Men | 0.841 | 0.813 | .001 | 0.817 | 0.793 | .003 | 0.833 | 0.822 | .015 |
Women | 0.812 | 0.786 | .002 | 0.801 | 0.786 | .015 | 0.796 | 0.815 | .003 |
All people | 0.803 | 0.794 | .015 | 0.787 | 0.813 | .002 | 0.807 | 0.799 | .015 |
aAUROC area under the receiver operating characteristic curve
bP value: P < .05 was considered to indicate statistical significance