Machine Learning and Statistics Analysis of Socioeconomic and Health Factors Impact on the Progress of Countries' Humanitarian Commitments

Abstract

Under-five Mortality Rate (U5MR), as one of the 17 Sustainable Development Goals established by United Nations, reveals the social commitment on children's health and international humanitarian development progress. In addition to traditional regression analysis and dimension-reduction factor analysis regarding the determinants of child mortality, this paper takes a step further and conducts cluster analysis using data mining and machine learning techniques with Python to better visualize and demonstrate the geospatial traits of global development progress on certain topic. The result of stepwise multivariate regression analysis suggests that the average life expectancy, female fertility rates and GDP per person of the area are the top three factors that affect U5MR. Factor analysis is then applied to reduce the variables into four dimensions, demographic factor, individual financial factor, national trade factor and Heath spending & Income factor. With the outcomes of the principal component analysis, Python is adopted to perform K-Means cluster analysis. Four classes, determined by elbow method and Silhouette experiment, are clustered to represent levels of development of countries. The results are visualized on a world map for intuitive interpretation. Supported and cross-verified by existing studies, sub-Saharan African countries require immediate attention and international assistance as the new-born and the mothers fall victims of inadequate fundamental, feasible and deliverable resources such as immunization, skilled attendant, early breastfeeding, and warmth. Through scientific and statistic methods, this paper is dedicated for international organizations, governments, and NGOs to optimize and facilitate recourses given the geospatial and unbalanced socioeconomic and health resources worldwide.