Social Inequalities and Quality of Life in Greater Cairo, Egypt: A Machine Learning Approach

Social inequality is a complex and multifaceted phenomenon and is extremely widespread globally. It is often influenced by historical, social, economic, cultural, environmental, and political conditions. It poses a wide variety of threats to the quality of life (QOL) of individuals and the welfare of society. Although social survey instruments like household questionnaires can detect detailed information about socio-economic inequalities and their context, it can be challenging to identify important predictors from a large number of explanatory variables with traditional statistical methods, such as regression analysis and correlation analysis. Machine learning (ML) techniques are well suited to pattern identification and can help identify important predictor variables from large complex social datasets. Based on the Egyptian Labor Market Panel Survey (ELMPS), this paper aims to apply a random forest classification algorithm to identify significant predictors with the greatest predictive power for household residence characteristics between different neighborhoods in Greater Cairo. The core goal is to examine the aspects of social inequalities and their impact on QOL and show how these factors change depending on the geo-spatial context. The results show that applying a random forest approach can identify nuances in predictors of different types of inequalities in different communities. Moreover, they confirm that factors foresting inequality are indissolubly linked to many aspects of QOL. Therefore, the random forest model and other machine learning techniques can be useful for improving the classification accuracy of social models and identifying patterns in complex social datasets.