Short-Term Forecasting of Dockless Bike-Sharing Demand with the Built Environment and Weather

Table 3

Result comparison of different models for bike-sharing demand forecasting.

Model	MAE	MAPE (%)	RMSE	R ^2	Time (s)
HA	8.375	38.80	12.421	0.183	1
MA	14.830	478.34	17.370	−0.598	1
ARIMA	9.870	55.79	13.954	0.593	4
Holt exponential smoothing	10.000	56.87	14.174	0.580	4
ANN	13.333	590.60	17.042	−0.539	18
LSTM	10.900	91.01	14.573	−0.125	656
GRU	9.526	70.47	12.264	0.383	727
SGCNPM	8.209	37.12	11.527	0.737	1320