Skip to main content

Table 2 Contingency table for the first feature set

From: Multivariate classification of urine metabolome profiles for breast cancer diagnosis

 

Actual

Output

Accuracy

Sensitivity

Specificity

  

Cancer

Normal

   

Decision Tree

      

Confidence=0.25

Cancer

47

3

94.00%

94.00%

94.00%

Pruning=true

Normal

3

47

   

Random Forest

      

Tree=500

Cancer

47

3

95.02%

94.00%

96.00%

Feature=6

Normal

2

48

   

Support Vector Machine (Linear)

      

Cost=1

Cancer

50

0

89.06%

100.00%

72.00%

Gamma=0.33

Normal

14

36

# of Support Vectors: 52

Support Vector Machine (Gaussian)

      

Cost=45

Cancer

49

1

95.16%

98.00%

92.00%

Gamma=0.33

Normal

4

46

# of Support Vectors: 22

Classification results for the first feature set (Table 1A)

  1. Contingency table showing number of cases classified for each of the diagnostic classes for the first feature set (Table 1A).
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us