Early prediction of sudden cardiac death risk with Nested LSTM based on electrocardiogram sequential features

Table 4 MAE results

Signals(.dat)	SVM(mV)	ESN(mV)	LSTM(mV)	Bi-LSTM(mV)	Nested LSTM(mV)
30	0.0671	0.0138	0.0076	0.0102	0.0052
31	0.0832	0.0192	0.0078	0.0087	0.0063
32	0.1041	0.0612	0.0134	0.0125	0.0098
33	0.1089	0.0278	0.0089	0.0091	0.0052
34	0.0871	0.0192	0.0081	0.0101	0.0072
35	0.1344	0.0906	0.0127	0.0118	0.0098
36	0.1237	0.0812	0.0169	0.0147	0.0072
37	0.1212	0.0923	0.0211	0.0209	0.0162
38	0.0971	0.0145	0.0079	0.0068	0.0049
39	0.2120	0.1514	0.0421	0.0437	0.0364
41	0.0389	0.0123	0.0051	0.0083	0.0033
43	0.0412	0.0282	0.0092	0.0069	0.0047
44	0.0441	0.0271	0.0065	0.0072	0.0039
45	0.0531	0.0122	0.0072	0.0083	0.0066
46	0.0631	0.0478	0.0122	0.0175	0.0110
47	0.0639	0.0323	0.0189	0.0171	0.0099
48	0.0331	0.0199	0.0068	0.0047	0.0058
50	0.0672	0.0352	0.0102	0.0143	0.0096
51	0.0931	0.0448	0.0223	0.0321	0.0182
52	0.0456	0.0381	0.0117	0.0161	0.0094

ISSN: 1472-6947