A Two-Stage Train Delay Prediction Method Based on Data Smoothing and Multimodel Fusion Using Asymmetry Features in Urban Rail Systems

Figure 4

Comparison of real and predicted values of different methods in the training set.