دراسة مقارنة بين طريقة الشبكات العصبية الاصطناعية و منهجية بوكس جانكينس للتنبؤ بمبيعات الكهرباء : دراسة تطبيقية في شركة "سونلغاز"

Abstract EN

In this work, we have made a case study which is to forecasting sales of Sonelgaz Company of low electricity tension directed to the families through collecting and recording the previous sales of Sonelgaz between the years 2010-2015 and then using them employed in the embodiment of two models, one econometric represented in Box– Jenkins method and the other mathematical represented in Artificial Neural Networks Method and then we have differentiated between them on the basis of statistical tests for the targeted model and on Theil Law and concluded that Box–Jenkins method presents the best and the most accurate model to get to close estimates to reality in terms of electricity sales of low tension belonging to "Sonelgaz" company.