Research Article

Short-Term Origin-Destination Based Metro Flow Prediction with Probabilistic Model Selection Approach

Table 4

The VAPE of inflow at each station on weekdays.

StationCombined ARMA NN SVR Averaging Adaboost kNN

11.262.043.291.391.271.301.56
21.393.661.63 1.411.391.401.31
30.732.644.031.060.730.770.88
40.964.931.681.350.961.011.19
50.843.841.371.320.850.870.86
60.903.971.051.070.911.011.29
70.873.971.05 0.910.890.881.21
83.155.124.485.193.153.324.99
91.601.782.032.731.641.562.7
100.872.832.921.070.870.911.62
110.692.580.990.770.690.741.72
120.722.791.170.920.720.730.83
130.863.181.771.080.870.911.17
141.104.891.81.721.111.151.34
150.822.183.151.060.830.861.01
160.833.011.351.060.840.890.89
170.983.881.781.620.981.011.03
182.848.394.826.082.853.033.66
191.592.31 1.581.871.591.581.94

# of best 13 0 0 0 5 1 0
Average1.213.582.211.771.221.261.64