A Simple Method of Residential Electricity Load Forecasting by Improved Bayesian Neural Networks

Table 11

Results on different datasets.

Home number∖Methods	IBNN (BNN_16)				TSNN (TS_BNN, time delays = 25)
	MSE	R	MAE	Time	MSE	R	MAE	Time
11	2.47e-2	0.87	0.10	17.48	2.38e-2	0.89	0.10	476.87
13	3.43e-2	0.85	0.09	12.05	3.24e-2	0.88	0.10	350.22
16	2.86e-2	0.87	0.10	7.79	2.83e-2	0.88	0.10	477.66
22	7.76e-2	0.72	0.16	9.73	7.24e-2	0.75	0.16	480.80
25	8.60e-2	0.90	0.15	11.94	8.34e-2	0.89	0.15	264.94
28	2.32e-2	0.87	0.07	21.89	2.68e-2	0.85	0.09	199.21
31	4.30e-2	0.83	0.09	10.56	4.78e-2	0.86	0.11	207.40
34	4.03e-2	0.82	0.11	12.84	4.25e-2	0.84	0.12	482.75
37	2.43e-2	0.89	0.09	12.86	1.81e-2	0.91	0.09	330.83
40	1.11e-2	0.91	0.06	14.64	1.12e-2	0.91	0.06	199.69
44	1.98e-2	0.79	0.07	11.09	2.11e-2	0.84	0.08	263.15
47	2.34e-2	0.90	0.09	9.64	2.36e-2	0.89	0.09	143.94
50	4.67e-2	0.51	0.13	7.49	4.58e-2	0.54	0.13	222.25
55	4.63e-2	0.89	0.16	5.63	4.83e-2	0.88	0.16	475.61
58	7.52e-2	0.84	0.17	11.77	8.09e-2	0.83	0.19	315.73