"Predicting and Factors Affecting Under-Five Age Mortality in Sub-Saharan Africa Using Machine Learning." Method and current data from health and demographic surveys (2018–2024)

The number of children that pass away before turning five out of every 1,000 live births in a population is known as under-five age mortality. Despite a rise in health measures, the Sub-Saharan region has a persistently high under-5 age mortality rate. Because of the intricate interactions between socioeconomic, health, and environmental factors, under-five mortality is still a major public health concern in the area. This study intends to predict and analyze the factors affecting under-five age mortality in Sub-Saharan Africa using machine learning approaches, leveraging recent demographic health survey data from 2018 to 2024. By utilizing ML techniques, the study hopes to uncover previously unidentified or under-explored risk factors that contribute to U5M. Finding important variables is essential for successful health interventions given the region's ongoing child mortality problem. Adaptive Boosting (AdaBoost), KNeighbors Classifier, Random Forest, Logistic Regression, and Artificial Neural Networks (ANN) were among the machine learning techniques we used. Based on our findings, the Random Forest algorithm had the highest accuracy (97.64%), followed by ANN (95.77%). The extensive dataset used in this study included 181,278 records from 16 Sub-Saharan African nations. Burkina Faso (2021), Cameroon (2018), Ivory Coast (2021), Gabon (2019–21), Gambia (2019–20), Ghana (2022), Guinea (2018), Kenya (2022), Liberia (2019–20), Madagascar (2021), Senegal (2019), Sierra Leone (2019), Nigeria (2018), Rwanda (2019–20), and Tanzania (2022) are the first, second, third, and fourth countries.