LogCAD: An Efficient and Robust Model for Log-Based Conformal Anomaly Detection

Table 6

Effectiveness of multiple models on HDFS log data set.

Injection rate (%)	Classifier	Accuracy	Precision	Recall	F1	MCC
5	CP-AB-LR	0.998	0.950	1	0.973	0.976
	CP-AB-SVM	1	0.992	1	0.996	0.996
	CP-AB-NB	1	0.992	1	0.996	0.996
	LogCAD	1	0.992	1	0.996	0.996
10	CP-AB-LR	0.998	0.96	1	0.98	0.976
	CP-AB-SVM	0.999	0.996	0.992	0.994	0.994
	CP-AB-NB	0.999	0.974	0.997	0.985	0.984
	LogCAD	0.999	0.996	0.992	0.994	0.994
15	CP-AB-LR	1	0.983	0.989	0.986	0.983
	CP-AB-SVM	0.999	0.989	0.987	0.988	0.986
	CP-AB-NB	0.999	0.972	0.991	0.982	0.981
	LogCAD	1	0.992	0.992	0.992	0.992
20	CP-AB-LR	0.999	0.942	0.982	0.963	0.960
	CP-AB-SVM	0.998	0.947	0.995	0.971	0.969
	CP-AB-NB	1	0.958	1	0.979	0.975
	LogCAD	0.999	0.975	0.996	0.985	0.985