Table 5 Feature weights calculated by information gain applied on 5-classes-dataset of fGOS.

From: Prognosis prediction in traumatic brain injury patients using machine learning algorithms

Row

Feature

Weight

1

GCSM0

0.1144

2

pupil

0.1037

3

age

0.0702

4

Marshall

0.0598

5

Cisterns

0.0580

6

DC

0.0383

7

INR

0.0335

8

1st BS

0.0261

9

IVH

0.0248

10

Shift

0.0237

  1. GCSM0: motor component of GCS on admission, DC: decompressive craniotomy, INR: international normalized ratio, BS: blood sugar, IVH: intraventricular hemorrhage.
Back to article page