Skip to main content

Table 1 Selected top 6 features in the first stage

From: Establishment of a differential diagnosis method and an online prediction platform for AOSD and sepsis based on gradient boosting decision trees algorithm

Order

Features

Shap-Importance

1

White blood cell count

0.048411

2

Arthralgia

0.039684

3

Monocyte percentage

0.035879

4

α1 acid glycoprotein

0.028332

5

Ferritin

0.025614

6

Sore throat

0.024988

7

Creatinine

0.022643

8

Lactate dehydrogenase

0.020723

9

Percentage of eosinophils

0.019778

10

APTT

0.017245

  1. APTT activated partial thromboplastin time