A Modified Gray Wolf Optimizer-Based Negative Selection Algorithm for Network Anomaly Detection

Table 5

Summary of results.

Dataset	Algorithm	Recall (%)	Accuracy (%)	Precision (%)	F1 (%)
NSL-KDD	NB	92.30	92.60	98.80	95.44
	KNN	91.30	92.60	99.00	94.01
	SVM	77.20	81.10	99.20	86.30
	DT	91.20	92.80	99.00	94.94
	RF	91.00	92.70	99.00	94.83
	DNN	91.30	92.50	98.90	94.90
	CNN	95.73	95.54	95.36	95.54
	LSTM	90.79	94.26	99.05	94.74
	RNSA	95.31	80.97	70.86	81.29
	HD-NSA	95.38	96.49	96.21	95.79
	MDGWO-NSA-(without DP-SUMIC)	94.76	95.31	98.97	96.82
	MDGWO-NSA	99.23	97.61	96.68	97.94
UNSW-NB15	NB	65.00	74.35	96.04	77.53
	KNN	96.00	93.71	94.00	94.99
	SVM	83.71	74.80	70.53	76.56
	DT	98.00	94.20	93.00	95.43
	RF	97.00	95.43	96.00	96.50
	DNN	67.80	75.03	90.30	77.45
	CNN	92.28	82.13	96.16	93.68
	LSTM	92.76	82.40	96.37	94.53
	RNSA	80.25	93.85	95.71	94.77
	HD-NSA	95.91	93.76	96.42	96.16
	MDGWO-NSA-(without DP-SUMIC)	93.57	94.46	95.18	94.37
	MDGWO-NSA	98.46	95.76	94.57	96.48
CICIDS-2017	NB	80.00	82.00	82.00	80.99
	KNN	94.56	92.33	94.55	94.55
	SVM	84.00	84.00	81.50	82.73
	DT	95.00	92.70	95.00	95.00
	RF	94.00	92.60	94.70	94.35
	DNN	95.03	92.97	95.03	95.03
	CNN	86.30	99.50	80.80	83.46
	LSTM	92.95	96.83	98.31	95.41
	RNSA	87.29	95.33	82.57	84.86
	HD-NSA	92.79	97.34	97.11	94.90
	MDGWO-NSA- (without DP-SUMIC)	92.53	94.38	96.75	94.59
	MDGWO-NSA	95.73	98.60	99.24	97.45