USING MACHINE LEARNING ALGORITHMS TO PREDICT ANEMIA IN CHILDREN UNDER 5 YEARS: A COMPARATIVE STUDY

Abstract

Anemia is a common health issue in young children, which can lead to serious consequences if not detected and treated early. In this study, we explore the potential of using machine learning algorithms to predict anemia in children under the age of five. We collected data from Kanti Children Hospital in Nepal, consisting of 700 data records, and selected six different machine learning algorithms for verification and validation, including Random Forest, Decision Tree, Naïve Bayes, Artificial Neural Network, Support Vector Machine, and Logistic Regression. The data was preprocessed, normalized, and balanced, and the algorithms were applied to improve accuracy in predicting anemia. We also applied ensemble learning methods, including Voting, Stacking, Bagging, and Boosting, to further improve performance. Our study found that Random Forest was the best performer with an accuracy of 98.4%. Feature analysis indicated that selecting the best features also contributed to improving accuracy. Balanced data was used to further validate the results. Our study highlights the potential of machine learning in predicting and preventing diseases in the field of health informatics, particularly in the case of anemia in young children