The Prediction of Seasonal ARIMA By using Exponential Smoothing Methods with Application

The reconciliation one of the time series model with ARIMA (P, D, Q) S seasonal to the average humidly in Mosul (1971-2001) have been done in this study.
Prediction It is reached to the best Models in the proportional ARIMA (2, 1, 4)8, because it is the less value for statistical standard AIC (k), MSE.
In current study, it reached also to the reconciliation of seasonal time series models after performance of introduction for series by making Programs (MATLAB) for it. The method adopted is Winters' Three Parameters Exponential (Smoothing) method, method of Holt-Winters Multiplicative and Holt-Winters Additive.
The third method choosed the best model of prediction for every method depending on the same statistic.
Finally, in this paper the best model is reconciliated for the introduction methods after performance, the difference for every method and the choose of the best prediction of every method depending on the same statistical methods.