Forecasting Algeria's General Industrial Index: a Comprehensive Data-Driven Analysis and Predictive Insights

EasyChair Preprint no. 11567

Abstract

The manufacturing sector serves as the cornerstone of a nation's economic, productive, and social influence. Given its substantial share of global energy consumption, embracing data-driven strategies represents a pivotal approach to informed decision-making. Machine learning techniques, renowned for their capacity to address intricate challenges, are increasingly favored for energy forecasting. This research introduces and assesses four analytical models - namely, Linear Regression, Exponential Smoothing, ARIMA, and SARIMA - in the context of predicting Algeria's General Industrial Index. In this study, these four models are applied to assess and predict Algeria's general industrial production for the forthcoming decade (2021–2030) by utilizing a 50-years dataset provided by The National Office of Statistics (NOS). The findings of this study reveal that the four forecasting methods provides a unique perspective on the future trajectory of the General Industrial Index in Algeria. Linear Regression is optimistic, Exponential Smoothing suggests stable growth, ARIMA predicts a decline, and SARIMA reinforces the declining trend with seasonality considerations. This paper significantly contributes to the literature by providing a comprehensive analysis of Algeria's general industrial production using four distinct forecasting models. The findings offer valuable insights for policymakers and industry stakeholders, facilitating informed decisions to sustain and enhance manufacturing output in Algeria.

Keyphrases: ARIMA, Data Analytics, Exponential Smoothing, Forecasting, Index of Industrial Production, linear regression, Manufacturing Industries, SARIMA

@Booklet{EasyChair:11567,
  author = {Siham Bouguern},
  title = {Forecasting Algeria's General Industrial Index: a Comprehensive Data-Driven Analysis and Predictive Insights},
  howpublished = {EasyChair Preprint no. 11567},

  year = {EasyChair, 2023}}