Abstract

Preterm birth is one of the major contributors to worldwide mortality and Nigeria has the 3rd highest number of infant mortalities in Africa. Complications arising from pre-term birth contribute to the deaths of more than 1 million infants each year. The victims of this menace are susceptible to some medical complications due to their low immune system. This is because they are not fully developed or have reached the mature gestation age of 37 to 40 weeks. Therefore, detecting women who are at risk of preterm birth is necessary through an evidence-based intervention model with an accurate computational approach to tackle it for the sustainability of the human race. This study therefore focused on advancing the prediction of preterm birth using machine learning. The data used for this study were sourced from Federal Teaching Hospital, Lokoja, Nigeria, as well as online sources to argument the data. Five supervised machine learning (ML) models which included Artificial Neural Network (ANN), Decision Tree (DT), k-nearest Neighbor (kNN), Support Vector Machine (SVM), and Naive Bayes were employed, evaluated and analyzed for preterm birth. The models were compared with some previous studies in this domain based on their performance and accuracy. ANN and DT models achieved the best results on the dataset compared to the other models.