A Multifeatures Spatial-Temporal-Based Neural Network Model for Truck Flow Prediction

Table 5

The RMSEs, MAPEs, and MAEs of various models.

Model		Evaluation	S1	S2	S3	S4	S5	S6	Mean
M-BiCNNGRU	Multifeatures/spatial/temporal	RMSE	26.46	18.71	12.82	3.53	26.09	14.74	17.06
		MAPE	7.62	5.50	7.25	6.84	9.71	5.93	7.14
		MAE	21.96	12.74	9.41	2.83	22.28	10.62	13.31
BiCNNGRU	Spatial/temporal	RMSE	42.04	27.05	13.88	8.79	28.82	24.02	24.10
		MAPE	8.42	9.95	9.92	12.61	9.16	10.72	10.13
		MAE	33.58	21.19	11.69	7.25	24.51	19.71	19.65
M-BiGRU	Multifeatures/temporal	RMSE	36.66	26.70	16.87	6.43	23.20	22.88	22.12
		MAPE	9.08	12.17	15.03	13.40	9.59	13.41	12.11
		MAE	28.43	18.24	13.84	5.08	18.30	17.72	16.94
Bi-GRU	Temporal	RMSE	46.60	23.76	14.54	9.03	30.13	24.55	24.77
		MAPE	11.47	15.84	15.51	18.67	10.19	13.89	14.26
		MAE	38.20	18.80	11.25	7.85	23.11	18.39	19.60
SVR	Temporal	RMSE	58.59	49.99	27.18	25.91	51.23	34.04	41.16
		MAPE	14.88	21.18	22.14	28.92	15.55	23.15	20.97
		MAE	25.02	22.54	19.19	28.10	33.14	19.29	24.55
ARIMA	Temporal	RMSE	66.43	48.89	30.00	15.38	50.03	41.74	42.08
		MAPE	15.43	20.78	21.46	29.28	16.39	18.33	20.28
		MAE	53.20	34.88	23.22	11.32	40.96	29.28	32.14