Defending Malicious Script Attacks Using Machine Learning Classifiers

Table 4

Results obtained from four classifiers for 80% training and 20% testing.

Classifier	Accuracy	TP rate	FP rate	Precision	Recall	-measure	ROC	Class
Naïve Bayes	95.06%	0.795	0.007	0.971	0.795	0.874	0.964	Malicious
		0.993	0.205	0.946	0.993	0.969	0.965	Nonmalicious
J48	99.22%	0.964	0	1	0.964	0.982	0.983	Malicious
		1	0.036	0.99	1	0.995	0.983	Nonmalicious
SVM	94.55%	0.88	0.036	0.869	0.88	0.874	0.922	Malicious
		0.964	0.12	0.967	0.964	0.965	0.922	Nonmalicious
KNN	97.14%	0.904	0.01	0.962	0.904	0.932	0.957	Malicious
		0.99	0.096	0.974	0.99	0.982	0.957	Nonmalicious