A thesis on utilizing machine learning models to predict material hardship

Abstract

This thesis addresses the concern of material hardship, defined by a paucity in resources necessary to fulfil fundamental needs, especially affecting children and families. Drawing insights from studies on the determinants of material hardship, including low income, unemployment, single parenthood, and financial literacy, the research employs a comprehensive methodology. It incorporates the integration of machine learning techniques to enhance the predictive capacity of identifying at-risk populations. The methodology advocates for a holistic approach, incorporating the transformative potential of machine learning techniques such as Logistic Regression, Non-Linear Support Vector Machine Model, Decision Tree etc. This paper highlights the transformative potential of machine learning in proficiently analyzing extensive datasets to recognize complex patterns. The positive correlation established between higher financial literacy, bill paying tendency, management of finances and improved economic outcomes stresses the potential impact of targeted financial education initiatives. Moreover, the research emphasizes the need for a proactive stance, advocating for the development of predictive models using historical evidence to anticipate and address material hardship in a timely manner. The inferences emphasize the necessity for targeted interventions and proactive measures, promoting social equity, resilience, and contributing to broader poverty reduction strategies