Predicting Child Mortality using Ensemble Machine Learning Models

V.Tejaswi, P.Harshitha, Noorunnisha, N.Sindhil Raj, Dr Marada Srinivasa Rao
Department of Information Technology, Maharaj Vijayaram Gajapathi Raj College, Vizianagaram, AndhraPradesh, India

Abstract—Child mortality, which means the death of children under the age of five, is a serious global issue, especially in low and middle income countries, where proper healthcare and nutrition are not easily available. Early prediction of the risk can help doctors and healthcare workers to take necessary actions and save lives. In this work, multiple machine learning models including traditional classifiers, tree-based algorithms, boosting techniques, and ensemble methods are implemented and compared. A custom ensemble model is developed by combining the outputs of multiple models . To make the system practical and user-friendly, the trained model is deployed using a flask-based application. The proposed system not only improves the prediction accuracy but also enhances the usability and interpretability. Present work is carried out during the period of eight months(Aug 2025 to Mar 2026). In this work, we focus on developing a predictive system using machine learning models to estimate the risk of child mortality and also it highlights which features influence child mortality the most. Keywords–Child Mortality, Machine Learning, Ensemble Model, Healthcare, Risk Score, Flask.