Table 3 Summary of experiments performed.

From: White blood cell image analysis for infection detection based on virtual hexagonal trellis (VHT) by using deep learning

Exp #

Graft Net

VHT- FE

Fused Features

Classifiers

Acc %

Sen %

Sp %

PPV %

NPV %

Error %

1

100

100

200

Cubic SVM

99.99

99.98

99.99

99.97

99.98

0.001

CG. SVM

99.92

99.95

99.97

99.99

99.98

0.001

Med. SVM

99.87

99.72

100

100

99.91

0.0013

2

200

200

400

Cubic SVM

99.98

99.99

99.98

99.97

100

0.01

CG. SVM

99.72

99.76

100

100

99.92

0.037

Med. SVM

98.38

98.36

99.93

99.80

99.45

0.162

3

250

250

500

Cubic SVM

99.98

99.99

99.99

99.99

99.99

0.01

CG. SVM

98.99

98.96

99.97

99.92

99.65

0.0101

Med. SVM

96.75

96.76

99.81

99.42

98.92

0.0325

4

300

300

600

Cubic SVM

99.98

100

100

100

100

0.0002

CG. SVM

99.00

98.68

99.97

99.92

9956

0.0100

Med. SVM

94.60

93.71

99.12

97.26

97.92

0.0540

5

400

320

720

Cubic SVM

99.82

99.76

100

100

100

0.0016

CG. SVM

97.83

97.44

99.92

99.75

99.15

0.0227

Med. SVM

88.68

89.87

95.25

86.37

96.56

0.1132

6

500

320

820

Cubic SVM

99.62

99.72

100

100

99.91

0.0038

CG. SVM

97.23

97.36

99.84

99.51

99.12

0.0277

Med. SVM

81.30

88.03

86.53

68.62

95.57

0.187

7

600

320

920

Cubic SVM

99.33

99.48

99.97

99.92

99.83

0.0067

CG. SVM

96.35

96.52

99.76

99.26

98.84

0.0365

Med. SVM

70.58

85.58

74.45

52.86

93.91

0.2942

8

750

320

1070

Cubic SVM

98.90

98.68

99.96

99.88

99.56

0.0110

CG. SVM

95.72

95.27

99.56

98.63

98.44

0.0428

Med. SVM

59.32

85.82

60.60

42.17

92.74

0.4068