Parameter | Iteration 0 | Iteration 1 | Iteration 2 | Iteration 3 | Iteration 4 |
Cutoff value | 1.20 | 1.27 | 1.26 | 1.25 | 1.24 |
NH | 145 | 149 | 142 | 141 | 141 |
NL | 131 | 127 | 134 | 135 | 135 |
Acc-WC | 0.985 | 0.960 | 0.949 | 0.945 | 0.949 |
Spec-WC | 0.992 | 0.946 | 0.962 | 0.962 | 0.962 |
Sens-WC | 0.979 | 0.972 | 0.937 | 0.931 | 0.938 |
Acc-Iter | 0.960 | 0.975 | 0.996 | 1.000 | |
Spec-Iter | 0.946 | 1.000 | 1.000 | 1.000 | |
Sens-Iter | 0.972 | 0.953 | 0.992 | 1.000 |
At each iteration, accuracy, specificity, and sensitivity were calculated with respect to initial classification at iteration 0 (Acc-WC, Spec-WC, and Sens-WC) and with respect to preceding iteration (Acc-Iter, Spec-Iter, and Sens-Iter). NH = number of subjects in AβH group; NL = number of subjects in AβL group.