Skip to Main Content
Table 6

The detailed prediction errors of each prediction method

YearProposed approach (108 m3)ARMA (108 m3)SVM (108 m3)LSSVM (108 m3)BP neural network (108 m3)Elman neural network (108 m3)
2013 −0.3791 1.8727 0.7138 1.3615 1.4503 −1.0998 
2014 −0.3585 0.7316 −1.4407 1.3967 0.2205 0.4027 
2015 −0.4886 −1.1808 1.0020 0.0307 0.1959 −0.4975 
2016 −0.3852 −1.0187 0.7167 −0.5391 −0.5584 1.1471 
2017 0.7010 0.8065 −1.4712 0.4788 −0.3070 0.9390 
2018 −0.4623 0.9796 −0.5191 −0.5972 −1.4073 −1.5616 
YearProposed approach (108 m3)ARMA (108 m3)SVM (108 m3)LSSVM (108 m3)BP neural network (108 m3)Elman neural network (108 m3)
2013 −0.3791 1.8727 0.7138 1.3615 1.4503 −1.0998 
2014 −0.3585 0.7316 −1.4407 1.3967 0.2205 0.4027 
2015 −0.4886 −1.1808 1.0020 0.0307 0.1959 −0.4975 
2016 −0.3852 −1.0187 0.7167 −0.5391 −0.5584 1.1471 
2017 0.7010 0.8065 −1.4712 0.4788 −0.3070 0.9390 
2018 −0.4623 0.9796 −0.5191 −0.5972 −1.4073 −1.5616 
Close Modal

or Create an Account

Close Modal
Close Modal