A Novel Machine Learning Model for the Detection of Epilepsy and Epileptic Seizures Using Electroencephalographic Signals Based on Chaos and Fractal Theories

Table 5

Comparison of classification accuracy (%) using the KNN classifier and PML.

Experiments	U-Sub	KNN (all_feat)	PML (sel_feat)
Exp 1	Z-N	96.7	100
	Z-F	99.4
	Z-NF	98
	O-N	99.1
	O-F	99.5
	O-NF	99.8
	ZO-N	99.13
	ZO-F	98.46
	ZO-NF	98.6
Exp 2	Z-S	100	100
	O-S	99.9
	ZO-S	100
Exp 3	N-S	98.6	100
	F-S	97.1	99
	NF-S	97.73	99.5
Exp 4	ZN-S	99.26	100
	ZF-S	98.2	98
	ZNF-S	98.25	99
	ON-S	99.33	100
	OF-S	98	98
	ONF-S	98.05	99.4
	ZON-S	99.5	100
	ZOF-S	98.65	99.4
	ZONF-S	98.76	99.6
Exp 5	Z-NS	99.8	100
	Z-FS	98.9
	Z-NFS	98.5
	O-NS	99.46
	O-FS	99
	O-NFS	98.8
	ZO-NS	99.65
	ZO-FS	99.05
	ZO-NFS	98.96
Exp 6	Z-N-S	99.26	100
	Z-F-S	96.2	99.2
	Z-NF-S	96.65	99.6
	O-N-S	98.4	100
	O-F-S	97.4	99.25
	O-NF-S	97.3	99.62
	ZO-N-S	99.25	100
	ZO-F-S	97.5	99.42
	ZO-NF-S	97.28	99.54