Table 2 Performance of predictive models of chronic cough in subpopulations with a higher prevalence of chronic cough.

From: Machine learning to identify chronic cough from administrative claims data

 

Full sample

Individuals with cough diagnoses

Females

Individuals ≥ 65 years of age

ROC-AUC

XG Boost

0.916

0.817

0.914

0.898

Logistic regression

0.907

0.823

0.902

0.882

Neural network

0.838

0.684

0.837

0.829

PR-AUC

XG Boost

0.229

0.103

0.225

0.288

Logistic regression

0.205

0.241

0.207

0.244

Neural network

0.135

0.055

0.138

0.167

  1. PR-AUC precision-recall, area under the curve; ROC-AUC receiver-operating characteristic, area under the curve.
Back to article page