An Ensemble Model of Machine Learning Algorithms for the Severity of Sickle Cell Disease (Scd) Among Paediatrics
Abstract
Balogun Jeremiah Ademola, Aderounmu Temilade, Egejuru Ngozi Chidozie and Idowu, Peter Adebayo
This study was motivated at developing an ensemble of 3 supervised machine learning algorithms for the assessment of the severity of sickle cell disease (SCD) among paediatric patients. The study collected data from a tertiary hospital in south-western Nigeria following the identification of variables required for assessing the severity of SCD. The study also adopted the use of 3 supervised machine learning algorithms namely: naïve Bayes (NB), C4.5 decision trees (DT) and support vector machines (SVM) for creating the ensemble model using a 10-fold cross validation technique. The models were created by adopting the algorithms in isolation and in combination of 2 and 3 which were compared. The developed models were evaluated in order to present the model with the best performance. The results of the study showed that using an ensemble of DT and NB alone provided the best performance. The study has implications in presenting a model for improving the assessment of the severity of SCD among paediatric patients in Nigeria.
PDFShare this article