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

Table 2

Predicted accuracy of different methods.

	Training set		Testing set
	MAE	RSME	MAE	RSME
NN	7.80	6.29	16.08	25.86
SVM	9.02	11.78	16.34	24.46
RF	9.02	11.78	15.40	25.44
Baseline	8.61	9.92	15.94	25.25
Averaged	7.42	9.32	15.36	23.86
Weighted	6.74	8.63	15.22	23.84
KNN	5.02	6.24	12.08	23.14