Defending Malicious Script Attacks Using Machine Learning Classifiers

Table 3

Results obtained from four classifiers by performing 10-fold cross-validation.

Classifier	Accuracy	TP rate	FP rate	Precision	Recall	-measure	ROC	Class
Naïve Bayes	97.99%	0.868	0.001	0.994	0.868	0.927	0.963	Malicious
		0.999	0.132	0.966	0.999	0.982	0.936	Nonmalicious
J48	98.64%	0.951	0.004	0.985	0.951	0.968	0.971	Malicious
		0.996	0.049	0.987	0.996	0.991	0.971	Nonmalicious
SVM	95.42%	0.848	0.017	0.93	0.848	0.887	0.916	Malicious
		0.983	0.152	0.96	0.983	0.971	0.916	Nonmalicious
KNN	96.41%	0.927	0.026	0.907	0.927	0.917	0.956	Malicious
		0.974	0.073	0.98	0.974	0.977	0.956	Nonmalicious