Review on EEG-Based Authentication Technology

Table 3

Shallow classification methods for EEG authentication.

Researchers	Tasks	Feature extraction	Classification	Accuracy (%)
Kong et al. [65]	MI	Node degree of brain network	LDA	99.1
	DRI mental task			99.3
Salem et al. [41]	MANHOB-HCI VEP	CNN	SVM	99.99
Seha et al. [52]	Listening	CCA	LDA	96.46
Wu et al. [34]	FRSVP VEP	Fisher LDA and logistic regression	HDCA	91.46
Koike-Akino et al. [35]	ERP	PCA and partial least squares	LDA	96.70
Brigham et al. [47]	Imagined speech	AR	SVM	99.76
Jayarathne et al. [66]	Listening + VEP + ERP	CSP	LDA	96.97
Keshishzadeh et al. [51]	Resting state	AR	SVM	97.43
Thomas et al. [31]	Resting state	Individual alpha frequency (IAF) delta band power (DBP)	Cross-correlation values and Mahalanobis distance	90
Bashar et al. [42]	Resting state	MSD, WPES	ECOC-SVM	94.44
Gui et al. [17]	VEP	WPD	ANN	90
Pham et al. [46]	MI	AR, PSD	SVDD	99.90
Zeynali et al. [67]	Mental task	DFT, DWT, AR	BN	85.97
			SVM	84.49
Wu et al. [34]	RSVP	Fisher LDA	HDCA with genetic algorithm	94.26