Concurrent, Performance-Based Methodology for Increasing the Accuracy and Certainty of Short-Term Neural Prediction Systems

Table 8

Traffic flow forecasting performances of different approaches.

Method	Error type
	MAE	RMSE	MSE	MAPE
Single ANN	20.46	23.39	547.34	266.59
Combined ANN
Averaged	16.60	21.32	454.45	186.23
MAE weighted	16.53	21.02	441.86	186.41
RMSE weighted	16.44	20.54	421.84	204.37
MSE weighted	18.17	22.31	497.88	187.39
Combined ANN-LIN.REG.
MAE weighted	10.21	13.00	168.67	102.84
RMSE weighted	10.22	12.39	153.41	98.40
MSE weighted	9.67	11.35	128.84	105.95
ARIMA	47.20	24.00	575.83	389.60