Effects of Data Augmentation with the BNNSMOTE Algorithm in Seizure Detection Using 1D-MobileNet

Table 4

Results of the 1D-MobileNet model in each patient, on accuracy, and on F1-score.

Patient	Accuracy (%)						F1-score (%)
	Before	ROS	SMOTE	BLSMOTE	SVMSMOTE	BNNSMOTE	Before	ROS	SMOTE	BLSMOTE	SVMSMOTE	BNNSMOTE
chb01	99.70	99.73	99.12	99.70	99.41	99.81	94.45	95.44	85.08	95.18	90.63	96.22
chb02	99.39	99.36	99.56	99.68	99.45	99.81	89.10	88.44	93.19	94.71	90.46	95.08
chb03	99.53	99.60	98.76	99.52	99.49	99.77	89.96	92.28	82.14	92.24	90.74	94.14
chb05	98.28	99.23	98.43	98.93	99.07	99.51	92.04	92.56	83.77	90.69	91.18	94.33
chb07	99.55	99.59	98.67	99.39	99.21	99.75	90.75	93.64	82.78	91.32	88.56	95.00
chb08	98.72	98.58	96.61	98.70	97.64	99.03	90.24	90.75	71.82	88.27	85.97	94.01
chb09	99.36	99.68	99.91	99.89	99.69	99.93	96.30	94.75	97.28	96.92	95.08	97.62
chb10	99.49	99.52	98.67	99.53	99.39	99.62	89.12	89.79	78.05	90.16	87.59	93.43
chb11	98.64	97.21	96.85	98.08	96.19	98.67	87.02	88.87	85.05	89.32	85.29	93.10
chb13	98.73	98.69	95.48	97.56	95.86	99.00	75.81	79.23	79.29	80.29	75.24	86.04
chb14	98.35	99.28	98.78	98.56	98.38	99.54	80.24	78.22	82.41	83.33	85.00	90.41
chb15	96.98	97.48	93.41	95.53	93.66	97.56	73.41	75.86	69.31	68.43	73.66	82.00
chb17	98.22	98.53	94.48	98.64	98.53	99.23	72.06	83.65	64.03	81.48	78.75	89.42
chb18	99.67	99.67	98.89	99.58	99.25	99.73	89.88	91.93	75.51	85.96	83.25	91.93
chb19	99.70	99.58	99.47	99.77	99.41	99.84	89.42	92.51	91.75	94.36	90.07	94.76
chb20	98.11	99.43	99.24	99.33	99.24	99.56	88.25	90.91	88.10	91.59	88.75	94.62
chb21	98.98	99.25	96.41	98.80	98.77	99.24	78.04	80.28	62.79	64.52	72.04	85.07
chb22	99.44	99.46	99.17	99.51	99.38	99.77	88.34	88.81	82.92	87.52	86.99	94.68
chb23	99.20	99.29	98.05	98.76	98.71	99.29	78.78	85.06	73.11	78.04	76.17	89.06
chb24	98.99	99.05	97.21	98.74	98.83	99.25	80.86	82.43	71.31	75.31	70.32	87.00
Average	98.95	99.11	97.86	98.91	98.48	99.40	85.70	87.77	79.98	85.98	84.29	91.90