Artificial Neural Networks for Analysis of Factors Affecting Birth Weight

Creative Commons License

Kirişci M. , Yılmaz H.

American Journal of Information Science and Computer Engineering, cilt.5, sa.1, ss.17-24, 2019 (Diğer Kurumların Hakemli Dergileri)

  • Cilt numarası: 5 Konu: 1
  • Basım Tarihi: 2019
  • Dergi Adı: American Journal of Information Science and Computer Engineering
  • Sayfa Sayıları: ss.17-24


Objective: This work presents 2017 data on births in Beykoz according to a wide variety of characteristics. Factors affecting birth weight (BW) of the newborn were examined with Artificial Neural Network (ANN). Method: There are 19 independent variables and one dependent dichotomous variable that form a database. These data were obtained from 223 samples. For this study, the analysis method used is ANN. Results: The variables as the gender of the baby, maternal age, Body Mass Index (BMI), nutrition habits, mother’s education, gestational age, maternal pre-pregnancy weight, maternal weight gain in pregnancy, mother's alcohol use, mother's cigarette use are considered. These variables have been found to have a significant effect on the BW. Conclusion: In this study, for the BW, a novel system based on ANN is proposed. When a person contacts a physician, it is very important for the physician to accurately estimate whether or not this person has a disease. This is because the physician's estimate is a principal factor in determining whether to discontinue treatment, to obtain more data by testing, or to treat the patient without exposing the patient to the risks of advanced diagnostic tests. It is believed that the prepared ANN system may be useful for physicians to make decisions about factors affecting the BW of babies during pregnancies. The results show that ANN system based on a learning method can assist in the prediction of BW.