Table 1 Performance of 3 claims-based predictive models of chronic cough.

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

Metric

XG Boost

Logistic regression

Neural network

Sensitivity (recall)

0.207

0.153

0.201

PPV (precision)

0.134

0.344

0.218

Specificity

0.974

0.996

0.989

NPV

0.985

0.987

0.988

Accuracy

0.960

0.983

0.977

ROC-AUC

0.916

0.907

0.838

PR-AUC

0.229

0.205

0.135

  1. NPV negative predictive value; PPV positive predictive value; PR-AUC precision-recall, area under the curve; ROC-AUC receiver-operating characteristic, area under the curve.
Back to article page