كفاءة طريقتي الشبكات العصبية وطريقة بوكس جنكنزفي التنبؤ مع حالات تطبيقية في العراق

Abstract

This paper attempts to make a comparison between  the most famous classic method of forecasting ; Box-Jenkins method , and the famous expirt forecasting method; neural  network method . The comparison is based on the forecast error that is evaluated for different types of time series data ; i.e. general and seasonal data . This study showed that the Neural method was very sutable for both data types with better results than box-jenkins method .hence different neural network models were fitted and the resulted forecast were compared with the results obtained from the (box-jenkins)  seasonal and non-seasonal  ARIMA models.the results showed superior performance of neural network technology; which makes it a valuable tool for decision making .