Prediksi Bayi Lahir Secara Prematur Dengan Menggunakan Metode C.45 Berbasis Particle Swarm Optimization Pada Klinik Umi

Abstract

Abstract - Babies born prematurely can occur when the pregnancy has not reached the mature gestational age. Pregnancy usually lasts about 40 weeks. Some risk factors for preterm birth include having given birth to a premature baby before and becoming pregnant with twins. The complications associated with preterm birth include immature lungs, difficulty regulating body temperature, difficulty eating, and slow weight gain. Premature babies can require longer or more intense nursery care, medication, and sometimes surgery. Until now, there have been several cases of preterm labor that have no known cause. There are a number of factors and health problems that can trigger premature labor, which are unhealthy mothers, smoking, a history of pregnancy, fetal conditions, psychological conditions. For this reason, the author intends to make a study on how to predict a patient who will deliver prematurely. In this study C4.5 algorithm model and C4.5 algorithm model based on particle swarm optimization are used to get the rule in predicting premature births of babies and provide accuracy values. more accurate. After testing with two models, namely C4.5 Algorithm and C4.5 Algorithm based on particle swarm optimization, the results obtained are C4.5 Algorithm produces an accuracy value of 94.30% and AUC value of 0.986 with a diagnosis level of Excellent Classification, but after The addition is C4.5 algorithm based on particle swarm optimization, the accuracy value is 97.91% and the AUC value is 0.997 with the diagnosis level of Excellent Classification. So that both methods have different levels of accuracy which is equal to 3.61%.

Keywords: gestational age, premature, C4.5 algorithm