CRND: An Unsupervised Learning Method to Detect Network Anomaly

Table 6

The evaluation results of CRND-200 and supervised method.

Model	Accuracy (%)	Precision (%)	Recall (%)	False-alarm (%)	F1-score
CRND-200	96.13	99.45	91.94	0.42	0.96
MLP-based	99.78	99.99	99.49	0.01	0.99
CRND-100000	99.59	99.83	99.26	0.14	0.99