A Hybrid Spatiotemporal Deep Learning Model for Short-Term Metro Passenger Flow Prediction

Table 6

The comparison of different models (outbound passenger flow).

Model	Terminal stations			Transfer stations			Regular stations
	RMSE	MAE	MAPE (%)	RMSE	MAE	MAPE (%)	RMSE	MAE	MAPE (%)
HSTDL	34.368	21.794	33.6	55.176	26.301	18.2	20.005	14.711	23.7
GBRT	39.778	24.563	41.0	81.581	45.176	23.2	27.521	18.077	29.0
CNN	64.037	30.055	43.5	91.401	46.753	31.8	39.956	22.364	39.1
LSTM	51.882	27.418	44.5	88.553	44.977	32.1	37.726	21.022	39.6
MLP	65.037	36.833	53.7	99.441	49.841	35.1	53.071	28.111	43.8
ARIMA	79.998	54.011	64.9	127.112	57.051	38.9	63.339	39.126	49.5