Golden Cub Predictor

The birth weight of an infant may come as a surprise at the end of a pregnancy. The birth weight is an important factor in the health of the mother and the delivery procedure of the infant. If the parents know the birthweight in advance, they may be more likely to schedule an early induction or opt for a Cesarean delivery to avoid complications that can result from larger babies, such as shoulder dystocia. In addition, a higher baby weight may allow the parents to skip purchasing certain baby items which are only needed for very small infants. For example, one can imagine a world in which your baby weight is predicted, and the Amazon delivers the appropriate size diapers after birth to your door, along with right sized clothing. This project will develop a baby weight predictor using information that is easily available from electronic health records from the parents, such as age, weight, and height of the parents. It can provide an initial prediction at the beginning of pregnancy, and then can continue to provide more accurate predictions throughout the pregnancy, with the collection of additional data.

The challenges for this project include securing biological paternal data, since this has a significant impact on infant birth weight; however, preliminary search of datasets only include maternal data. Another challenge is the unpredictability of birth timing. For example, even if a model accurately predicts the baby’s weight from week to week, if the mother goes into labor two weeks early, the birth weight would be significantly reduced. So an accurate model must have accurate predictions for both weight and timing of birth.