Automated seizure diagnosis system based on feature extraction and channel selection using EEG signals

Brain Informatics

Table 6 Evaluation of the metric parameters for the seven classifiers of the random case testing method

	SVM	Ensemble	KNN	LDA	Logistic regression	Decision tree	Naïve Bayes
TP	39	42	37	41	40	40	11
TN	29	24	31	22	22	27	37
FP	8	13	6	15	15	10	0
FN	4	1	6	2	3	3	32
Accuracy	85	82.5	85	78.75	77.5	83.75	60
Sensitivity	90.69767	97.67442	86.04651	95.34884	93.02326	93.02326	25.58
Specificity	78.37838	64.86486	83.78378	59.45946	59.45946	72.97297	100
PositivePre	82.97872	76.36364	86.04651	73.21429	72.72727	80	100
F1	86.66667	85.71429	86.04651	82.82828	81.63265	86.02151	40.75
FallOut	21.62162	35.13514	16.21622	40.54054	40.54054	27.02703	0
MisRate	9.302326	2.325581	13.95349	4.651163	6.976744	6.976744	74.42