ST-LSTM: A Deep Learning Approach Combined Spatio-Temporal Features for Short-Term Forecast in Rail Transit

Table 1

Performance of four models.

Model	ME	MAE	RMSE	MRE	Operation time
SARIMA	360.70	51.33	82.37	29.63%	75 min
PSO-SVR	153.45	18.89	27.48	21.46%	8 min
LSTM	183.59	22.82	32.29	26.45%	37 min
ST-LSTM	145.72	18.42	25.89	20.77%	38 min