Ensemble Deep Learning for Biomedical Time Series Classification

Table 6

Statistical results of different classification models.

Model	Sp (%)	Se (%)	GMean (%)	value	Acc (%)	AUC	NPV = 95%		value
							TPR (%)	FPR (%)
Explicit [6]	83.16 ± 4.20	81.66 ± 4.20	82.32 ± 1.42	0.0039	83.40 ± 1.68	0.8993 ± 0.02	22.74 ± 33.90	2.95 ± 4.40	0.0156
Explicit	84.16 ± 4.28	82.06 ± 4.27	83.02 ± 1.44	0.1641	84.16 ± 1.76	0.9073 ± 0.01	30.53 ± 36.86	3.96 ± 4.78	0.0156
Implicit	88.29 ± 3.00	78.15 ± 4.30	83.01 ± 2.00	0.0039	84.68 ± 1.96	0.9079 ± 0.02	25.16 ± 37.82	3.26 ± 4.90	0.0078
Fusion	86.86 ± 3.51	80.23 ± 4.49	83.40 ± 1.80	—	84.84 ± 1.82	0.9117 ± 0.02	36.11 ± 34.34	4.30 ± 4.74	—

The classification models are obtained by “explicit training” and “implicit training,” respectively, and the results are based on subview prediction.