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

Table 9

Comparison between different classifiers with Haar, dB4, and Sym 8 wavelet features without feature selection.

Wavelets	Classifiers	Sensitivity	Specificity	Accuracy	F1 score	Error rate	G-mean
Haar	Linear regression	57	66	61.5	59.68586	38.5	61.58507
	Nonlinear regression	58	57	57.5	57.71144	42.5	57.5007
	GMM	68	75	71.5	70.46632	28.5	71.58989
	K-NN	55	87	71	65.47619	29	73.01289
	SVM (linear)	62	72	67	65.26316	33	67.14976
	SVM (polynomial)	65	62	63.5	64.03941	36.5	63.51087
	SVM (RBF)	69	85	77	75	23	77.58279
dB4	Linear regression	70	57	63.5	65.7277	36.5	63.70707
	Nonlinear regression	55	66	60.5	58.20106	39.5	60.61733
	GMM	63	84	73.5	70.39106	26.5	74.40527
	K-NN	61	57	59	59.80392	41	59.01332
	SVM (linear)	72	57	64.5	66.97674	35.5	64.79557
	SVM (polynomial)	63	53	58	60	42	58.07519
	SVM (RBF)	53	55	54	53.53535	46	54.00154
Sym8	Linear regression	54	64	59	56.84211	41	59.08392
	Nonlinear regression	59	87	73	68.60465	27	74.63016
	GMM	55	73	64	60.43956	36	64.41616
	K-NN	55	81	68	63.21839	32	69.12302
	SVM (linear)	71	57	64	66.35514	36	64.24879
	SVM (polynomial)	57	83	70	65.51724	30	71.23203
	SVM(RBF)	90	95	92.5	92.30769	7.5	92.46621