Neural Networks for Predicting Technological Use in Neonatal Care

Health care providers are increasingly concerned about the rising costs of the Neonate Intensive Care Units (NICU) and therefore a model that accurately predicts the technological use can be potentially beneficial for health care planning. This paper concerns the assessment of neural networks for predicting the use of GASOMETRY in a Brazilian NICU. Our results show that neural networks may not be superior to multiple linear regression models when no clear non-linear relationship exists.