Table 2 Comparison of performance metrics across the base pipelines.
From: Machine learning models incorporating genotype and ancestry improve severe asthma risk prediction
Metric | SNP base pipelines | LA base pipelines | ||
|---|---|---|---|---|
\(P1_{SNP}\) | \(P2_{SNP}\) | \(P1_{LA}\) | \(P2_{LA}\) | |
AUC | 0.678 ± 0.091 | 0.641 ± 0.072 | 0.602 ± 0.103 | 0.605 ± 0.108 |
Weighted precision | 0.647 ± 0.062 | 0.643 ± 0.071 | 0.591 ± 0.103 | 0.578 ± 0.089 |
Weighted recall | 0.642 ± 0.059 | 0.635 ± 0.066 | 0.586 ± 0.099 | 0.575 ± 0.084 |
Weighted specificity | 0.638 ± 0.062 | 0.631 ± 0.064 | 0.589 ± 0.100 | 0.573 ± 0.088 |
Balanced accuracy | 0.640 ± 0.060 | 0.633 ± 0.064 | 0.588 ± 0.100 | 0.574 ± 0.086 |
