Support Vector Regression-Based Recursive Ensemble Methodology for Confidence Interval Estimation in Blood Pressure Measurements

Table 4

Evaluating of the SVREM algorithm using the BHS and AAMI protocols using the results of maximum amplitude algorithm (MAA), neural network (NN), support vector regression (SVR), deep neural network (DNN), , and SVREM on () measurements, where is the AdaBoost with DNN model.

Methods	SBP mean absolute difference (%)			DBP mean absolute difference (%)			SBP/DBP BHS	SBP AAMI	DBP
	≤5 mmHg	≤10 mmHg	≤15 mmHg	≤5 mmHg	≤10 mmHg	≤15 mmHg	Grade	ME (SDE)	ME (SDE)
MAA	47.29	84.94	95.53	56.00	86.82	96.94	C/B	0.15 (9.56)	-0.17 (8.10)
NN	53.65	85.88	95.29	65.88	94.12	98.58	B/A	-0.31 (7.84)	0.47 (6.92)
SVR	62.59	86.12	95.53	74.12	93.65	96.94	A/A	0.10 (7.15)	-0.34 (6.45)
	71.06	90.82	95.53	81.18	96.24	99.29	A/A	-0.06 (5.82)	0.14 (4.98)
SVREM	69.17	87.06	95.53	76.94	94.12	98.12	A/A	-0.12 (6.29)	0.97 (5.33)