Skip to main content

Table 2 The accuracy, FPR, TPR and G-Mean of each classifier when using the best mean settings (i

From: Detecting unknown malicious code by applying classification techniques on OpCode patterns

Classifier

Best averaged settings

Classifier optimal settings

Difference in accuracy

 

Accuracy

TPR

FPR

G-Means

Optimal settings

Accuracy

 

RF

95.146

0.843

0.020

0.909

6-gram; TF; GR; top300

95.375

0.229

BDT

94.436

0.850

0.031

0.908

2-gram; TF; FS; top300

94.649

0.213

DT

93.009

0.835

0.045

0.893

2-gram; TF; FS; top300

93.741

0.732

LR

92.965

0.819

0.041

0.886

2-gram; TF; DF; top300

92.965

0.000

ANN

92.136

0.785

0.043

0.867

3-gram; TF; DF; top300

92.199

0.063

SVM

92.136

0.785

0.043

0.867

3-gram; TF; DF; top300

92.199

0.063

NB

84.537

0.517

0.068

0.694

5-gram; TF; FS; top50

85.802

1.265

BNB

84.537

0.517

0.068

0.694

2-gram; TF; DF; top50

86.683

2.146

  1. The table also depicts the optimal settings of each classifier and the difference in accuracy with the averaged settings.