Table 3.
Cross-validation | Acca | 95% CI | NIRb | P value (Acc > NIR) | Kappa | Sensc | Specd | PPVe | NPVf | F1g | Prevalence | Detection rate | Detection prevalence | Balanced accuracy |
4-fold cross- validation |
0.96 | (0.96- 0.97) |
0.61 | <.001 | 0.92 | 0.94 | 0.98 | 0.96 | 0.96 | 0.95 | 0.39 | 0.36 | 0.38 | 0.96 |
10-fold cross- validation |
0.97 | (0.96- 0.97) |
0.61 | <.001 | 0.93 | 0.95 | 0.98 | 0.96 | 0.97 | 0.96 | 0.39 | 0.37 | 0.38 | 0.97 |
aAcc: accuracy.
bNIR: no-information rate.
cSens: sensitivity.
dSpec: specificity.
ePPV: positive predictive value.
fNPV: negative predictive value.
gF1: F1 score.