Skip to main content
. 2020 Jul 29;7:101015. doi: 10.1016/j.mex.2020.101015

Table 3.

RMSEs and MAPEs for all methods considering the training and the testing data of the Java-Bali load series in 2016.

Method RMSE
MAPE (%)
Training Testing Training Testing
SSA-LRF 703.72 447.30 2.83 1.74
SSA-LRF-WFTS Chen's 247.55 179.44 0.96 0.69
SSA-LRF-WFTS Yu's 255.27 189.57 1.02 0.75
SSA-LRF-WFTS Cheng's 226.94 172.03 0.87 0.63
SSA-LRF-WFTS Lee's 266.94 211.28 1.08 0.81
SSA-LRF- NN 18-10-1 152.71 129.90 0.55 0.50
TLSAR* 343.83 385.98 1.29 1.41
TBATS 225.29 293.72 0.77 0.96
DSHW 151.85 191.00 0.53 0.69
ARIMA(5,0,4)-NN(31,16-1) 81.22 1770.05 0.28 6.61
TLSNN(31-16-1) 93.13 366.06 032 1.43

period: 168, number of harmonics: 21, ARIMA(4,1,4).