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

Table 12

Electric load consumption forecasting performances of different approaches.

Method	Error type
	MAE	RMSE	MSE	MAPE	Maximal error
Single ANN	34.42	42.31	1789.95	39.00	93.45
Combined ANN
Averaged	30.67	38.18	1457.48	39.66	73.15
MAE weighted	30.55	38.05	1448.13	39.57	72.93
RMSE weighted	30.53	38.00	1443.75	38.43	72.85
MSE weighted	29.72	37.28	1389.59	39.52	72.49
Combined ANN-LIN.REG.
MAE weighted	35.49	41.59	1729.46	50.78	87.10
RMSE weighted	35.83	41.94	1759.12	51.42	87.45
MSE weighted	37.58	43.49	1897.06	54.26	87.63
SARIMA	28.27	35.13	1234.13	42.38	85.18