Table 1. Results of different methods on the dataset PEMSBAY.

Classification	Methods	PEMSBAY (10/15/20 min)
		MAE	MAPE (%)	RMSE
Classical methods	HA	2.41/2.48/2.56	5.62/5.78/5.97	5.37/5.58/5.76
	SVR	1.43/1.60/1.74	3.46/3.87/4.26	3.23/3.60/3.94
	RF	1.35/1.54/1.72	2.90/3.42/3.90	2.69/3.15/3.56
	KNN	1.45/1.62/1.78	3.20/3.65/4.08	2.92/3.34/3.72
State-of-the-Art methods	STGCN	1.39/1.68/2.02	3.24/4.02/5.11	2.50/3.10/3.67
	FC-GAGA	1.36/1.65/1.83	3.08/3.91/4.44	2.69/3.45/3.97
	AGCRN	1.37/1.47/1.58	3.23/3.50/3.76	2.77/3.03/3.27
Our methods	GST-HCN	1.32/1.51/1.68	3.17/3.71/4.21	2.41/2.82/3.18
	HST-HCN	1.41/1.60/1.76	3.68/4.18/4.64	2.61/3.01/3.31
	LST-HCN	1.25/1.46/1.64	2.81/3.45/3.96	2.27/2.73/3.11
	ST-HCN	1.22/1.40/1.57	2.73/3.23/3.73	2.19/2.58/2.94

Notes.

The best experimental results for each setting are bolded.