Abstract:
A new born’s health is a primary factor that determines the overall health of a human being and its life expectancy. Therefore, its health should be monitored not only after birth but also when the baby is still growing in the womb. Birth weight is one of the crucial aspects to be observed. Low birth weight is among the main problems that new borns face. Low birth weight (LBW) is the weight at birth less than 2500g as defined by the World Health Organization. A global estimate of 15 to 20 percent of total live births are low birth weight representing over twenty million births every year. In Kenya, the rate of children born with low weight is 8 percent. Several methods have been used to measure and approximate birth weight in clinical practice including obstetric ultrasound, symphysio-fundal height measurements and abdominal palpation. However, these methods are associated with reliability and accuracy challenges therefore, calling for more robust methods. This research aimed at creating a machine learning model for predicting low birth weight using the maternal risk factors that have been found to be associated with low birth weight. Secondary data from the 2014 Kenya Demographic Health Survey was utilized where the variables were extracted from the births recode file. The study
population included mothers between the age of 15 to 49 years. The machine learning
algorithms employed were logistic regression, decision trees, random forest, support vector machines, gradient boosting and xtreme gradient boosting. Using performance evaluation metrics namely; accuracy, precision, recall, F1 score, and ROC-AUC, the random forest model was found out to be the most robust with 0.956679 accuracy, 0.956831 precision, 0.956679 recall an F1 score of 0.95666 and an AUC of 0.988. In addition, variable importance was performed using the random forest approach to ascertain the maternal risk factors that are the most important to predict low birth weight. It was found out that mother’s weight was the most important variable for predicting low birth weight. The other important variables found were; mothers height, mother’s age and the number of antenatal visits attended by the mother during pregnancy. Machine learning techniques are increasingly being used to provide information to guide health policy. This research merits further modelling, research and more consultation.
