A Spatial-Temporal Self-Attention Network (STSAN) for Location Prediction

Table 2

Parameters setting on four datasets.

				H	d
New York	1.0	1.0	40	8	512
Tokyo	1.0	1.0	20	32	768
Los Angeles	0.8	0.8	50	2	704
Houston	0.8	0.9	60	1	640