Skip to main content

Table 2 A comparison among source separation algorithms with respect to different classifiers

From: A novel approach to emotion recognition using local subset feature selection and modified Dempster-Shafer theory

 

Runica

SOBI

COMBI

JADE

p-value

Index channels

14

16

17

15

–

MLP

 Accuracy (%)

77.16

79.57

76.33

80.28

0.0646

 Time (min)

118.46

120.78

116.89

113.45

KNN

 Accuracy (%)

79.11

81.46

77.16

73.28

0.0894

 Time (min)

112.56

110.32

118.96

103.52

Bayes

 Accuracy (%)

82.57

84.65

78.24

79.67

0.0743

 Time (min)

121.32

122.85

119.65

118.45

SVM

 Accuracy (%)

84.65

86.78

85.96

83.13

0.0531

 Time (min)

115.43

112.47

108.75

111.65

Modified DST

 Accuracy (%)

88.49

90.54

86.72

89.32

0.0417

 Time (min)

122.25

120.82

123.67

126.95

p-value

0.0631

0.0301

0.0472

0.0787

Â