Using The Hybrid Model SARIMA - ELMAN (ERNN) To Predict The Monthly Maximum Temperature Of Baghdad

Abstract

The process of strategic planning of a particular phenomenon and the process of making appropriate decisions depend on the process of accurate future prediction through the improvement of a model to represent that phenomenon , This research uses Seasonal Time Series Hybrid Models assuming that the series include components linear , which can be described as the The Multiplicative Sasonal Model SARIMA (p, d, q) (P, D, Q) S and non-linear Component , describing the The Model Elman Recurrent Of The Neural Network Model For the monthly rates of maximum temperatures for the city of Baghdad for the years (1937 - 2013) Using a set of statistical measures, it has been found that the hybrid model SARIMA (P, D, q) (P, D, Q) S-ELMAN has small values for Mean Absolute Error (MAE) and Mean Absolute Deviation (MAD) and Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) And Standard Error of Prediction (SEP),and high value for Therefore, it was used to predict the monthly average temperature for Baghdad City for the years 2017-2018.