LCHI: Low-Order Correlation and High-Order Interaction Integrated Model Oriented to Network Intrusion Detection

Table 5

Average detection performance (%) of different methods for multiclassification on the AWID, NSL-KDD, UNSW-NB15, CICIDS 2017, CICIDS 2018, and DAPT 2020 datasets.

Dataset	Model	ACC	F1	FNR	FPR
AWID	SVM	96.33	97.70	1.92	0.27
	BN	94.88	87.25	1.96	2.67
	RF	96.13	72.71	40.02	0.00
	CNN	96.51	78.79	28.69	0.97
	LSTM	97.13	83.74	27.90	0.001
	DNN	98.54	91.61	15.47	0.002
	MCA	96.10	80.56	24.75	0.36
	LCHI	99.60	99.68	1.02	0.005
NSL-KDD	SVM	56.90	66.95	39.29	27.79
	BN	61.80	78.13	34.73	2.39
	RF	76.05	75.69	37.68	3.09
	CNN	77.42	84.57	22.90	6.90
	LSTM	77.72	83.70	23.30	8.70
	DNN	81.51	85.35	21.50	7.20
	MCA	77.76	82.38	27.34	4.93
	LCHI	83.97	89.40	15.40	6.16
UNSW-NB15	SVM	69.70	80.90	17.10	27.00
	BN	57.06	76.59	26.12	23.34
	RF	72.29	82.12	16.25	24.78
	CNN	75.71	83.04	16.70	21.10
	LSTM	76.20	83.54	17.20	18.90
	DNN	77.04	88.57	14.32	9.53
	MCA	76.14	87.68	17.05	7.66
	LCHI	80.78	92.97	8.00	7.25
CICIDS 2017	SVM	98.41	98.17	2.05	1.23
	BN	92.22	90.33	16.25	1.29
	RF	99.16	99.03	1.92	0.005
	CNN	97.78	97.37	4.92	0.16
	LSTM	98.07	97.73	4.40	0.04
	DNN	99.53	99.46	0.04	0.80
	MCA	98.52	98.70	0.43	2.85
	LCHI	99.78	99.81	0.28	0.13
CICIDS 2018	SVM	96.60	91.86	14.78	0.09
	BN	96.15	92.13	0.00	4.96
	RF	97.10	93.15	12.82	0.00
	CNN	97.29	93.59	12.04	0.00
	LSTM	95.14	87.90	21.59	0.00
	DNN	98.44	96.40	6.95	0.00
	MCA	97.38	93.55	11.97	0.05
	LCHI	99.61	99.14	1.55	0.05
DAPT 2020	SVM	97.88	97.85	1.93	2.28
	BN	88.42	87.44	22.16	0.20
	RF	98.10	98.09	0.14	3.59
	CNN	96.48	96.34	5.33	1.78
	LSTM	96.74	97.16	3.68	1.88
	DNN	98.47	98.46	0.39	2.62
	MCA	96.29	97.12	2.01	3.79
	LCHI	99.11	99.12	1.17	0.60