Unveiling Global Economic Stratification: A Machine Learning Framework for Multi-Dimensional Macroeconomic Analysis

Saad Saadouni, Siham Ammari and Souad Habbani

Data and Metadata, 2025, vol. 4, 1180

Abstract: Introduction: Traditional econometric approaches to multi-country macroeconomic analysis face critical limitations in capturing complex, non-linear relationships across diverse economic systems. Objective: This study aims to introduce a comprehensive machine learning framework, implemented in Python, that transcends conventional VAR model constraints by analyzing 13 key macroeconomic indicators across 217 countries (2010–2025). Method: Advanced clustering techniques (K-means) and ensemble learning (Random Forest), along with Principal Component Analysis (PCA), were applied to reveal hidden economic stratification patterns previously undetectable through traditional methods. Result: The analysis uncovers four distinct global economic clusters representing differentiated development trajectories, with middle-income economies comprising the majority of observations (57.4%). Fiscal indicators demonstrate exceptional forecasting accuracy through Random Forest algorithms, while growth dynamics remain inherently unpredictable, revealing fundamental asymmetries in economic system behaviors. Conclusions: This study demonstrates that machine learning techniques, implemented in Python, can systematically identify which macroeconomic relationships are structurally determined versus stochastically driven. This differential predictability framework provides immediate policy implications for targeted intervention strategies, enabling policymakers to focus resources on controllable fiscal mechanisms rather than pursuing futile attempts to predict volatile growth patterns.

Date: 2025
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:dbk:datame:v:4:y:2025:i::p:1180:id:1056294dm20251180

DOI: 10.56294/dm20251180

Access Statistics for this article

More articles in Data and Metadata from AG Editor
Bibliographic data for series maintained by Javier Gonzalez-Argote ().