Skip to main content

Table 2 Algorithms comparison of computation time, target feature selection and predictive performance

From: Artificial Intelligence based wrapper for high dimensional feature selection

Model

p

n

Performance

Target feature selection (F1 Score)

Predictive performance (RMSE)

Computation time (minutes)

StW

IGIS

mIGIS

AIWrap

StW

IGIS

mIGIS

AIWrap

StW

IGIS

mIGIS

AIWrap

1_M

50

50

0.48

0.00

0.18

0.47

0.55

1.14

1.14

0.43

7.57

 < 0.02

0.07

24.85

2_M

50

100

0.71

0.00

0.95

0.63

0.29

1.11

0.30

0.29

10.68

 < 0.02

0.60

23.93

3_M

100

75

0.29

0.00

0.75

0.33

0.64

1.19

0.81

0.48

11.30

 < 0.02

0.62

18.22

4_M

100

100

0.42

0.18

0.18

0.43

0.36

1.18

1.18

0.36

31.52

 < 0.02

0.12

33.07

1_I

15

50

0.41

  

0.73

1.20

  

0.38

0.97

  

2.18

2_I

25

50

0.26

  

0.39

1.32

  

0.49

2.72

  

7.23

  1. Values in Bold means best results