Skip to main content

Table 9 Summary of results—subject 2 in Haxby dataset

From: Exploring stability-based voxel selection methods in MVPA using cognitive neuroimaging data: a comprehensive study

Method

Mean accuracy (%)

STD (%)

Average number of selected features

Stability

SVM-MI

70.83

13.41

123

0.57

SVM-Corr

71.88

12.07

357

0.86

SVM-RFE

78.13

14.23

195

0.67

LASSO (\(\varTheta\) = 0.6)

55.21

6.44

94

0.57

LASSO (\(\varTheta\) = 0.7)

48.96

17.24

90

0.52

LASSO (\(\varTheta\) = 0.8)

48.96

17.24

97

0.46

LASSO (\(\varTheta\) = 0.9)

43.75

12.50

104

0.46

ENet (\(\varTheta\) = 0.6)

32.29

16.39

269

0.64

ENet (\(\varTheta\) = 0.7)

50.00

18.46

264

0.62

ENet (\(\varTheta\) = 0.8)

53.13

22.06

214

0.58

ENet (\(\varTheta\) = 0.9)

47.92

12.87

252

0.55