Prognosticating Risk Factors of Undernutrition in Under Five Years Children in Nigeria Using Ensemble Technique

Abstract

Insufficient nutritional intake and nutrient imbalances is referred to as malnutrition, which signifies a universal health matter that has a substantial consequence on children, leading to short-term and long-term consequences for their growth, development and general comfort. This research utilized the potentials of stack ensemble techniques in enhancing the prediction of Child undernutrition in Nigeria. The study exploits three key indicators such as: underweight, stunting and wasting from Nigerian Multiple Indicator Cluster Survey (MICS) 2021 dataset. Four different machine learning classifiers were employed namely: decision tree (DT), random forest (RF), k-nearest neighbors (KNN) and the stack ensemble technique meta classifier that is concatenated with logistic regression (LR) to form stack ensemble technique in other to improve the prognostic risk factors of children under the age of five, considering their nutritional status. Based on their ability to predict outcomes, these classifiers were assessed and contrasted using standard machine learning evaluation metrics like. Accuracy, precision, recall and F1-score respectively. The effectiveness of machine learning techniques, specifically the Stack Ensemble technique, in prognosticating and comprehending the factors that donate undernutrition in Nigerian children is demonstrated by this study.  The results offer important information for developing policies and focused actions.