DOI: 10.11648/j.ijdst.20261203.12 ISSN: 2472-2235

Machine Learning Approach to Assessment of Nutritional Status of Children with Cardiac Disease in Ethiopia

Tayu Abebe, Abdisa Hunduma
Show PDF Cite
Nutritional status is an important determinant of health outcomes among children with cardiac disease, yet identifying children at risk of poor nutritional status remains challenging. Machine learning methods may provide useful tools for identifying important predictors and improving the classification of nutritional status in pediatric cardiac patients. The main objective of the study is to predict the nutritional status of children with cardiac disease using various machine learning methods, such as decision tree, logistic regression, random forest, support vector machine, and eXtreme Gradient Boosting. The predictive performance of each machine learning method was evaluated based on precision, accuracy, recall, f1 score, and the Kappa statistics value. The results indicate that the eXtreme Gradient Boosting method is recommended for predicting children nutritional status with precision, accuracy, recall, f1 score, and kappa statistics values: 0.7812, 0.7995, 0.7799, 0.7884, and 0.6299, respectively. The features that determine the nutritional status of children with cardiac disease are ROSS/NYHA, types of cardiac disease, pulmonary hypertension, anemia, pneumonia, and residence. This research will assist policy makers and healthcare providers to develop a framework for implementing necessary interventions and care practices to prevent complications associated with nutritional status in cardiac patients.

More from our Archive