Skip to main content

Table 2 Classification accuracy results for SVM, LR and CNN

From: Wavelet image scattering based glaucoma detection

Data

  

WISN

Maximum

Partitioning

Classification

Image

Parameter

F1-score

Scheme

Algorithm

Representation

Setting

(%)

Random

Support

Gray Scale

\(s = 125, q = [1, 1]\)#8

98

 

Vector

Red Channel

\(s = 100, q = [4, 1]\)#5

94

 

Machine

Green Channel

\(s = 75, q = [1, 1]\)#33

91

  

Blue Channel

\(s = 125, q = [3, 1]\)#10

82

 

Logistic

Green Channel

\(s = 125, q = [1, 1]\)#8

85

 

Regression

Gray Scale

\(s = 100, q = [3, 2]\)#4

82

  

Blue Channel

\(s = 125, q = [1, 1]\)#8

80

  

Red Channel

\(s = 125, q = [4, 1]\)#12

79

 

Convolutional

NA

Red Channel

82

 

Neural

   
 

Network

   

Hospital

Support

Gray Scale

\(s = 125, q = [1, 1]\)#8

89

 

Vector

Blue Channel

\(s = 125, q = [4, 1]\)#37

86

 

Machine

Red Channel

\(s = 100, q = [4, 1]\)#5

85

  

Green Channel

\(s = 125, q = [2, 1]\)#9

83

 

Logistic

Red Channel

\(s = 125, q = [3, 2]\)#11

82

 

Regression

Gray Scale

\(s = 125, q = [4, 3]\)#14

77

  

Green Channel

\(s = 150, q = [4, 2]\)#31

72

  

Blue Channel

\(s = 150, q = [3, 2]\)#29

67

 

Convolutional

NA

Blue Channel

83

 

Neural

   
 

Network

  Â