Determinants of Egypt's unemployment rate with machine learning algorithms

Abstract

This research endeavored to harness the capabilities of machine learning algorithms, specifically gradient boosting (GB) and random forest (RF), to ascertain the optimal algorithm for precisely analyzing the determinants influencing Egypt's unemployment rate. Among the algorithms employed, the study identified the gradient boosting (GB) algorithm as the most accurate in this context. The investigation discerned vital variables that substantially influence Egypt's unemployment rate. The analysis revealed that the industry sector value notably impacts the unemployment rate, accounting for 37.3%, and the GDP growth contributes 23.7%. Additionally, variables such as the value of imports (11.7%) and labor force participation (9.5%) emerged as significant determinants. Moreover, the research demonstrated that the unemployment rate exhibits negative correlations with factors including the value added in the agriculture sector, foreign direct investment (FDI), GDP growth, and gross fixed capital formation. Conversely, the unemployment rate positively correlates with variables such as inflation, human development index, imports, labor force participation, industry sector value added, and services sector value.

Keywords

• unemployment rate
• machine learning algorithms
• gradient boosting
• random forest