Skip to main content

Table 2 Model metrics from the validation cohort before and after resampling of the training data

From: The prognostic value of gastroesophageal reflux disorder in interstitial lung disease related hospitalizations

 

Original data model

Resampled data model

Sensitivity (%)

0%

73.43%

Specificity (%)

100%

66.15%

Precision

N/A

0.2688

Negative Predictive Value (%)

85.50%

93.62%

Accuracy (%)

85.50%

67.20%

F1 Score

0

0.4

MCC

N/A

0.3

AUC(ROC)

0.75

0.76

  1. F1 score = (2*precision*recall)/(precision + recall). MCC = TP*TN—FP*FN/sqrt((TP + FP)*(TP + FN)*(TN + FP)*(TN + FN))