Developing Deep Survival Model for Remaining Useful Life Estimation Based on Convolutional and Long Short-Term Memory Neural Networks

Table 3

Results of FD001 with two analyses and different TW.

	Loss function (regression)				Loss function (Weibull)
TW	MAE	RMSE	^2	Score	MAE	RMSE	^2	Score
20	17.26	23.13	0.65	3,737	20.24	28.64	0.84	12,492
30	15.06	19.06	0.73	1,065	15.08	21.50	0.91	2,324
40	12.81	17.89	0.79	994	14.14	19.91	0.91	2,164
50	11.29	14.90	0.86	431	13.14	19.23	0.91	5,861
60	10.20	14.04	0.88	310	13.69	20.36	0.90	31,427
70	10.80	14.33	0.87	254	13.98	20.01	0.90	5,888
80	10.32	14.14	0.86	198	11.63	15.45	0.93	325
90	10.36	14.15	0.87	200	10.50	13.98	0.94	231
100	9.99	13.98	0.79	190	11.17	15.02	0.91	254
110	9.91	13.81	0.77	183	11.03	15.43	0.90	389
120	9.44	13.85	0.86	161	9.79	13.73	0.93	206