Table 2 The parameter setting, sensitivity, and specificity of five classifiers in approach 1 and approach 2. Abbreviations: Sensitivity and specificity in approach 2 are expressed as mean ± standard deviation.

From: Exploratory study on classification of lung cancer subtypes through a combined K-nearest neighbor classifier in breathomics

Classifier

Approach 1

Approach 2

Parameters

Sensitivity (%)

Specificity (%)

Parameters

Sensitivity (%) (mean ± SD)

Specificity (%) (mean ± SD)

KNN

K = 28

100

0

K = 2; 1; 2

76.09 ± 9.96

41.67 ± 9.55

RF

N = 215

91.30

18.75

N = 295; 155; 265

81.88 ± 3.32

12.50 ± 6.25

SVM

C = 1

100

0

C = 18; 29; 24

60.87 ± 40.73

47.92 ± 29.54

 

gamma = 0.01

  

gamma = 0.3; 0.3; 0.3

  

MLP

M = 5

60.87

37.5

M = 17; 66; 70

60.87 ± 2.17

33.33 ± 3.61

PLS-DA

n = 2

91.30

12.5

n = 6; 8; 5

63.05 ± 3.76

43.75 ± 6.25

Back to article page