Skip to main content

Table 3 Number of selected features and samples, and comparison of classification accuracies obtained by different classifiers for the classification of multiclass PDT samples

From: SFSSClass: an integrated approach for miRNA based tumor classification

Classifier

After feature and sample selection

Classification accuracy (%)

 

No. of miRNAs

No. of Samples

No. of Classes

 

SFSSClass

91

37

9

82.35

USC

187

77

11

41.17

DFL(Discretized data)

3

23

4

58.82

C4.5

3

23

4

52.94

RIP

3

23

4

35.29

NB

42

23

4

35.29

kNN1

42

23

4

47.05

kNN5

42

23

4

58.82

SVM

42

23

4

58.82

Classifier Fusion

40(miRNA)+40(mRNA)

68

11

76.47

Bagged Fuzzy kNN

40

68

11

70.58

PNN

173

68

11

70.58