Predictive Models of Academic Success and its Link to Economic Mobility: A Comparative Study in Developing Countries

Abstract

This article explores the use of predictive models to estimate academic success and its relationship with economic mobility in developing country contexts. Through a comparative analysis between Colombia, India and Kenya, it examines how academic and socioeconomic variables are integrated into machine learning algorithms to identify patterns that influence both educational performance and social ascent. Quantitative methods based on logistic regression and decision tree models are employed, using data from national surveys and educational bases. The results show that factors such as parental schooling, school infrastructure, and access to technologies are significant predictors of academic performance and are positively correlated with intergenerational economic mobility. The policy implications for reducing social gaps by strengthening education systems are discussed.