Skip to main content
. 2023 Jun 29;14:1138489. doi: 10.3389/fimmu.2023.1138489

Table 2.

Summary of performance of model on training and test datasets.

Data AUC (95% CI) AP SBrS Spiegelhalter-z (p-value)
Validation1 0.81 (0.79–0.84) 0.58 0.27 0.34 (0.74)
Test2 0.81 (0.55–0.90) 0.68 0.30 −0.29 (0.77)

1Validation data refers to thirty leave-out parts in repeated CV framework. 2Test data is the 20% of the whole study subjects which was used to mimic an external data source for external validation.