Short-Term Inbound and Outbound Passenger Flow Prediction for New Metro Stations Based on Clustering and Deep Learning

Table 4

MAPE between the new stations and the existing metro stations.

Station	Residential class (%)	Office class (%)	Business class (%)	Comprehensive class (%)
Zhenlongxi station	28.32	24.08	25.17	15.46
Zhongxin station	13.21	28.41	29.31	23.74
Zhucun station	31.42	29.04	12.82	26.95
Kengbei station	9.94	24.32	25.71	21.72

The bold values show the importance of cluster prediction for different types of metro stations. The accuracy of passenger flow prediction based on the same type of metro station is obviously higher than other types.