Skip to main content

Table 2 Classification metrics across different models

From: Data mining polycystic ovary morphology in electronic medical record ultrasound reports

ModelAccuracyClass labelSensitivitySpecificityPPVNPV
Gradient boosted trees0.961 (0.9471, 0.9721)PCOM-absent0.99770.99820.99770.9982
  Unidentifiable0.94740.97050.93870.9748
  PCOM-present0.91460.97610.92590.9723
Rules-based classifier0.976 (0.9645, 0.9846)PCOM-absent0.98850.99820.99770.9912
  Unidentifiable0.9660.98520.96900.9838
  PCOM-present0.96680.98290.94720.9894
  1. PPV Positive predictive value; NPV Negative predictive value
\