Forecasting with ARIMA model on External trade of ALgeria

Abstract

Look into the future is a main branch of Statistics. So every government, public and private organizations, as well as an individual has always been interested into the future, particularly to well predicting and therefore to well planning in many domain. using theoretical
results, we are sure that the ARIMA model takes an important place in predicting problems leading to decision making (Bosq, 1998). In the economy of a country and in order to attain a better growth, modeling and forecasting is the most important tool now, this can be done by one of the statistical technique called a Time series analysis (Brockwell and Davis, 1996). The computer tools and specialized software as well R helps us exploit the theoretical results obtained in this analysis (Belguerna and Benaissa, 2014). In this paper we tried to show how building a ARIMA model for a time series with particular reference of Box and Jenkins approach on annually total External trade of Algeria from the year 1963 to the year 2014 with the free software called R. Validity of the fitted model is tested using standard statistical techniques (Farouki, 2014). The fitted model is then used to forecast some future values of Imports and export of Algeria, Using these techniques a confidence interval is then obtained.