Fig. 1. shows the comparison of four kinds of models about the total water requirement of the true and regression
value measured by billion cubic meters. Under the condition of making minimum mean square error(MSE), the first
model, i.e., the model of the total water requirement determined by the influencing factors of this year, makes the
best result of autoregression. Therefore, we select the first model to be our forecasting model. Then we obtain the
forecasting of influencing factors from the year of 2012 to 2016 via matlab toolbox about ARMA model. Finally, we
use the first model above to forecast the total water requirement measured by billion cubic meters from the year of
2012 to 2016 and the results are shown in Table 5.
Fig. 1. shows the comparison of four kinds of models about the total water requirement of the true and regression
value measured by billion cubic meters. Under the condition of making minimum mean square error(MSE), the first
model, i.e., the model of the total water requirement determined by the influencing factors of this year, makes the
best result of autoregression. Therefore, we select the first model to be our forecasting model. Then we obtain the
forecasting of influencing factors from the year of 2012 to 2016 via matlab toolbox about ARMA model. Finally, we
use the first model above to forecast the total water requirement measured by billion cubic meters from the year of
2012 to 2016 and the results are shown in Table 5.
การแปล กรุณารอสักครู่..