Towards the Development of a Cloud Computing Intrusion Detection Framework Using an Ensemble Hybrid Feature Selection Approach

Table 13

Comparison with related work.

Source	FS method (type)	ML algorithm (s)	CA	ER	PR	RC	FPR	FNR	FAR (FNR + FPR)/2
Proposed binary detection engine using HEFSM	Hybrid	HEFSM (DT, RF, KNN)	99.55	0.45	99.52	99.54	0.44	0.46	0.45
Luo and Xia [38]	Filter	SVM	94.36	—	—	—	—	—	—
Eesa et al. [40]	Wrapper	DT	91.99	—	—	91	3.917	—	—
Bhattacharya and Sel-vakumar [51]	Hybrid	BN, J48, KM	96.46	—	—	—	—	—	—
Osanaiye et al. [1]	Filter	DT	99.67	—	—	99.76	0.42	—	—
Ambusaidi et al. [37]	Filter	SVM	99.91	—	—	98.76	0.28	—	—
Belouch et al. [44]	Wrapper	REPTree	89.85	—	—	—	—	—	—
Moustafa and Slay [47]	Hybrid	EM, LR, NB	82.1	—	—	—	—	—	17.5
Mogal et al. [48]	Hybrid	NB, LR	99.82	—	—	—	—	—	—
Idhammad et al. [39]	Filter	ANN	99.2	—	—	—	—	—	0.02
Vijayanand et al. [45]	Wrapper	SVM	98.95	—	—	—	4.1	18.5	11.3
Aljawarneh et al. [46]	Wrapper	NB, J48, RT	99.81	—	—	—	—	—	—
Anwer et al. [50]	Hybrid	J48, NB	88	—	—	—	—	—	—
Sun et al. [41]	Wrapper	SVM	91.68	—	—	—	—	—	—
Pham et al. [42]	Wrapper	EC (J48)	84.25	—	—	—	2.79	—	—
Besharati et al. [43]	Wrapper	EC (DT, LDA, ANN)	97.51	—	—	—	—	—	—
Mohammadi et al. [49]	Hybrid	DT	95.03	—	—	95.23	1.65	—	—
Maza and Touahria [15]	Hybrid	NB, MLP, SVM, KNN, DT	99.11	—	—	—	—	—	—
Tama et al. [22]	Hybrid	RoF, CF	85.8	—	—	88	—	—	—

DT: Decision Tree; SVM: support vector machine; REPTree: reduced error pruning tree algorithm; EM: expectation-maximisation clustering; LR: logistic regression; BN: bayesian networks; NB: Naïve Bayes; KM: K means learning algorithm; ANN: artificial neural networks; J48: J48 decision tree; EC: ensemble classifiers, bagging, or boosting; CF: conjunctive rule; LDA: linear discriminant analysis; MLP: multilayer perceptron; RoF: rotation forest.