A comparative study of classification algorithms for risk prediction in pregnancy

Pregnancy is one of the most delicate stages in every woman´s life requiring increased medical care and attention. Pregnancy induces a variety of abnormalities which may lead to severe complications if unnoticed or neglected. Pregnancy complications are health problems that are caused due to changes in physiological parameters during the period of gestation. Pregnancy complications can lead to severe maternal illness by women during pregnancy, at delivery and after delivery. The aim of this paper is to predict the present complications in the health of a pregnant woman using two classification algorithms namely C4.5 decision tree classification algorithm and Naive Bayes Classification Algorithm. The selected algorithms are powerful and popular tools used for the tasks of classification and prediction in Data mining. These two algorithms use pregnancy data collected from pregnant women in different stages of pregnancy to predict their present health state and the associated health complications. This study focuses on identifying the best algorithm among the two classification algorithms to predict the health status of each pregnant woman and its associated complication. Applying these classification techniques on pregnancy related data the status of risk in the health of any pregnant women can be identified and maternal and fetal morality rate can be reduced.