Saihood, QusaySonuc, Emrullah2024-09-292024-09-292022978-3-030-97255-4978-3-030-97254-71865-09291865-0937https://doi.org/10.1007/978-3-030-97255-4_12https://hdl.handle.net/20.500.14619/37741st International Conference on Emerging Technology Trends in Internet of Things and Computing (TIOTC) -- JUN 06-08, 2021 -- ELECTR NETWORKAnemia is the most common disease among children under school age, especially in developing countries, due to a lack of understanding about its causes and preventive measures. In most cases, anemia refers to malnutrition and is closely related to demographic and social factors. Previously, statistical methods were used to predict anemia among children and identify associated factors. It was concluded that this is not a good way. Following the success of machine learning (ML) techniques in exploring knowledge from clinical data in healthcare, it was a good chance to explore the knowledge of social factors associated with childhood anemia. In this study, we compared the performance of eight different ML techniques for predicting anemia in children using social factors to find the most appropriate method. ML techniques achieved promising results in predicting and identifying factors associated with childhood anemia. Multilayer perceptron (MLP) has the best accuracy of 81.67% with all features, while Decision Tree (DT) has the best accuracy of 82.50% when we applied feature selection methods. The explored knowledge of the social factors associated with anemia can guide nutritional practices and factors essential to child health. Additionally, identified factors can help prevent anemia outbreaks for appropriate intervention by governments and healthcare organizations.eninfo:eu-repo/semantics/closedAccessMachine learningClassificationAnemia in childrenIron deficiencySocial factorsConference Object10.1007/978-3-030-97255-4_122-s2.0-85127679100181Q4167WOS:000790816200012N/A