A Framework on Performance Analysis of Mathematical Model-Based Classifiers in Detection of Epileptic Seizure from EEG Signals with Efficient Feature Selection

Table 11

Related researches on EEG signal identification and its performance measurement constraints.

Authors	Features	Classifier	Accuracy in %
Rajaguru and Prabhakar [4]	Discrete wavelet transform (Haar, dB4, Sym8)	SVD	97.3
Murugavel and Ramakrishnan [5]	Wavelet transform with approximate entropy	SVM with ELM	96
Truong et al. [6]	EEG features	Hills algorithm	Sensitivity: 91.95; Specificity: 94.05
Manjusha and Harikumar [7]	Detrend fluctuation analysis with power spectral density	K-means clustering and KNN	Sensitivity: 90.48; Specificity: 92.85
Radüntz et al. [8]	EEG features	SVM and ANN	95.85 and 94.04
Ghaemi et al. [11]	Improved binary gravitation search algorithm with wavelet domain features	SVM	80
Kumar et al. [37]	—	Improved Elman neural network	96
In this paper	Haar wavelet features	SVM (RBF)	77
	dB4 wavelet features	GMM	73.5
	Sym 8 wavelet features	SVM (RBF)	92.5
In this paper	Haar wavelet + PSO features	SVM (RBF)	87
	dB4 wavelet + PSO features	K-NN	84.5
	Sym 8 wavelet + PSO features	SVM (RBF)	98