Beam Journal of Economic Studies
Volume 9, Numéro 2, Pages 698-724
2025-09-15

A Hybrid Methodology Combining Statistical Learning And Econometrics: Forecasting Algeria’s Real Gdp Using Lasso-ols Algorithmic Approach



Authors : Bouziane Mohamed . Boudali Belkacem . Guelailia Redouane .

Abstract

This study explores the integration of statistical learning and OLS, intending to model the Real GDP over Algeria macroeconomic data. For this ending, we use sequential algorithm. At first step, algorithm extends the macroeconomic dataset by performing data transformation, and then it pools the dataset into multiple folds, each fold contains all dataset excluding one variable. Variable selection and OLS regression are iterated for each fold, subjecting resulting models to accuracy tests. Model with unsignificant feature(s) is banned from model selection, else the model exhibiting the lowest cross-validated RMSE is chosen for RGDP prediction. Despite model interpretability, our proposed methodology demonstrates its effectiveness for conducting accurate results. Consequently, the model performed to predict Algeria’s Real GDP achieves high-accuracy, consistently establishing 99.99% of goodness of fit and meeting several linear regression assumptions. The findings provide valuable insights into economic performance mechanisms, particularly within the Algerian context. These insights may also be generalizable to other cross-sectional datasets and applicable for modeling various macroeconomic aggregates.

Keywords

Statistical Learning ; Econometrics ; Data Processing ; LASSO ; Variable Selection ; Regression ; Real GDP

Les articles similaires

Measuring The Impact Of Digital Banking Services On The Financial Performance Of Banks: An Applied Case Study On The Banque Extérieure D’algérie (bea) Using Lasso And Ardl-lasso Methods

Bouatelli Mohamed . 
pages 153-170.

استخدام منهجية بوكس – جنكينز Box- Jenkins للتنبؤ بالناتج المحلي الإجمالي في السودان 2010- 2030م. Using Box- Jenkins Methodology To Forecasting Gdp In Sudan 2010-2030

محمد معتز . 
ص 119-133.

Economic Foresight As A Strategic Option To Achieve Diversification Of Gdp In Algeria, An Econometric Study To Anticipate The Diversification Index In Gdp (2020-2030) Using Ardl Model

Ghrieb Zaara .  Zeraoula Rafik . 
pages 65-81.

Forecasting Gdp For European Union (eu) Using An Arima Model

Latreche Walid .  Khallout Manal . 
pages 647-659.

صافي تدفقات رؤوس الأموال الأجنبية المباشرة وإجمالي الناتج المحلي في الجزائر: دراسة اقتصادية قياسية باستخدام نماذج الانحدار الذاتي للفجوات الزمنية المبطأة Ardl Net Foreign Direct Capital Inflows And Gdp In Algeria: An Econometric Study Using Ardl Methodology

مختاري عادل .  بن البار امحمد .  بوريش مهني . 
ص 260-275.

Formats de citation