Table 2 Performances of the Rel-En and AMM algorithms on the training and validation cohorts. The Rel-En algorithm performed better in the two cohorts both in terms of false negative and false positive rate
Rel-En | AMM | |
|---|---|---|
Training cohort | ||
False negative | 0.41% | 1.74% |
False positive | 0.27% | 2.40% |
Total error rate | 0.67% | 4.14% |
Validation cohort | ||
False negative | 0.76% | 2.65% |
False positive | 0.12% | 0.84% |
Total error rate | 0.88% | 3.49% |