Skip to main content

Table 6 The importance of features of prediction model

From: A study on predicting the length of hospital stay for Chinese patients with ischemic stroke based on the XGBoost algorithm

Ranking

Variables

Descriptions

Feature importance

1

Hemiplegia aphasia

Did the patient have symptoms of partial aphasia on admission

0.486

2

MRS

Ability of life (MRS) score at admission

0.305

3

NIHSS

NIHSS score at admission

0.221

4

TIA

Whether the patient had transient ischemic attack (TIA symptoms) on admission

0.181

5

Operation or not

Is the patient treated surgically

0.113

6

Coma index

Coma index

0.031

7

Critical or not

Was the patient critically ill at admission

0.027

8

Occupation

Different occupation types

0.021

9

Dizziness

Dizziness or not

0.018

10

Respiratory infection

Respiratory tract infection or not

0.017