Table 3 Variable ranking based on the mean rank of all models based on shapley additive explanations approach.

From: Machine learning for characterizing risk of type 2 diabetes mellitus in a rural Chinese population: the Henan Rural Cohort Study

Model

 

LR

CART

GBM

ANN

RF

SVM

Mean rank

Feature importance rank

Sweet flavor

3

2

1

4

1

3

2.33

Urine glucose

5

1

3

6

2

1

3

Age

2

4

2

5

4

2

3.17

Heart rate

8

10

4

10

6

8

7.67

Creatinine

7

13

6

9

9

6

8.33

Waist circumference

4

20

11

7

11

4

9.5

Uric acid

10

19

7

14

12

7

11.5

Pulse pressure

16

7

10

11

10

20

12.33

Insulin

12

8

14

15

18

13

13.33

Hypertension

15

32

9

18

5

11

15

  1. LR indicates logistic regression; CART, classification and regression tree; GBM, gradient boosting machine; ANN, artificial neural network; RF, Random forest; SVM, Support vector machine.
Back to article page