Comparing and Analyzing Applications of Intelligent Techniques in Cyberattack Detection

Table 7

Comparative study of the performance of different research citations.

References	Detection rate (%)	Accuracy (%)	Far	Technique	Datasets
[20]		98.70	95.60%	SVM, naïve Bayesian	MIT Lincoln Lab, IDS
[21]	96.11	—	24.88	PSO, SVM	KDDCUP99
[23]	—	—	—	AIS	nit DARPA LLDOS 1.0
[24]	98	—	—	MPSO	KDDCUP
[26]	98.53	—	0.0374	PSO, SVM	NSL-KDD
[27]	95	95	—	BARTD, cuckoo	KDDCUP
[28]	—	98.2/99.55	1.19/0.21	MLP, PSO, GSA, cuckoo	DARPA and WINTER
[72]	98.4/96	—	—	GA	DARPA and WINTER
[25]	96.7	—	—	PSO	KDDCUP
[69]	—	85.13	—	PSO	LLDOS
[83]	—	93.25	0.02	ABC	KDDCUP
[46]	—	99.4	—	PSO, SVM	KDDCUP
[32]	—	99.25	0.75	PSO, SVM	PMU2015
[55]	—	84.29	—	Multiple ELM	NSL-KDD
[71]	97.64	—	—	MSVM, PSO	KDDCUP
[21]	—	97.7	0.002	Hybrid PSO	KDDCUP
[77]	96.11	—	3.89	PSO-SVM	KDDCUP99
[74]	99.92	—	0.029	PSO RF	KDDCUP99
[84]	—	89.6	—	RNN	NSL-KDD
[84]		92		LSTM	NSL-KDD
[85]		93.20, 78.1, 66, 96.6		DNN	KDDCUP99, NSL-KDD, UNSW-NB15, WSN-DS
[86]	81.8	—	—	SVM RBF	KDDCUP