Table 2 Performance of SMFIF for classifying Sjögren’s disease versus controls

From: A multi-criterion feature integration framework for accurate diagnosis of Sjögren’s disease using routine laboratory tests

 

Testing set (3429 vs 4097)

Internal validation set (4203 vs 5126)

External validation set (139 vs 406)

ACC

0.858(0.850–0.866)

0.859 (0.852–0.866)

0.960(0.942–0.973)

Recall

0.832(0.821–0.845)

0.825 (0.813–0.833)

0.871(0.822–0.917)

F1-score

0.842(0.832–0.851)

0.841 (0.833–0.849)

0.917(0.883–0.947)

PPV

0.852(0.839–0.862)

0.857 (0.847–0.867)

0.968(0.929–0.992)

AUC

0.929(0.923–0.935)

0.934 (0.929–0.939)

0.964(0.943–0.986)

Specificity

0.879(0.869–0.888)

0.887 (0.880–0.897)

0.990(0.979–0.998)

  1. Data are metrics (95% CI).
  2. ACC accuracy, AUC area under the receiver-operating characteristic curve, PPV positive predicted value.
Back to article page