Application of machine learning approach to investigate the prevalence of stunting among under-five children in Ethiopia: A Case of Amhara and Afar Regions

In Ethiopia stunting is long-term year-round health problem, which makes difficult to end all forms of malnutrition by 2030. However, there has been progress in reducing the stunting level among children. Stunting is a global public health issue that is an essential determinant for the health of children. This research work aims to identify determinants of stunting and design a machine learning model for the prediction of under-five child stunting, which can support the decision-making process. Predicting the correct delivery method has significance in reducing malnutrition, especially the stunting rate by avoiding complications associated with wrong stunting prediction. Random Forest, Logistic Regression and Support Vector Machine classification algorithms are implemented as super learner to develop an under-five child stunting prediction model. Confusion matrix and its derivatives; Accuracy, Precision, Recall and F1-score are used to evaluate the performance of the proposed model. This study clarifies how a predictive model classifies the stunting condition using machine learning approach. The Support Vector Machine model shows the best accuracy with under-five child stunting. The study result shows that the model implemented using Support Vector Machine classification algorithms achieved the best accuracy and recall results of 74.0% and 73.61%, respectively.